NATIONAL MARITIME MUSEUM

Greenwich, London SEIO 9NF

THE LASHED-LUG OF THE EASTERN ARCHIPELAGOES

MARITIME MONOGRAPHS AND REPORTS

No. 54- 1982 The lashed-lug boat of the eastern archipelagoes,

the Alcina MS and the Lomblen whaling .

by

G. Adrian Horridge

Maritime Monographs and Reports

No. 54 - 1982

Published by the Trustees of the National Maritime Museum ISSN 0307-8590

ISBN 0 905555 61 9

© Crown copyright

Printed in England by Rank Xerox Copy Bureaux Ltd. ~,~------~1~ : __------r-C------~ JVO~------__ ___ : ~~: ___ ------. ~ ( ....__ N ' IOTEl TOBAGO 20°- . ~ .

~ ) '-, 9 123 Manil~a'• ... [) MINOORci\).9 ~ SOUTH CHINA ~ v Q'c:, r~:- ~ .~ :- r;;· V~ , SAMAR . ~ l ,,, E . P . P cit~u city - - lo• PA VISAY~0_-J \ f> 0I·• SEA MINDANAOI

CELEBES SEA

BORNEO (KALIMANTAN) ~ANGKA -• ~ ··0

JAVA SEA

ARAFURA SEA 10•- - 10° I ~ no• 130°

Figure 1

Map of the Eastern Archipelago with places mentioned.

ii OJNTENTS

List of Illustrations

Foreword

Acknowledgements

Introduction

The method of construction

The Alcina manuscript

The Manuscript

"On the construction capacity and variety of native boats". "How they continue, up to the launch". "On boat types, sizes and handling".

Other Early Accooots

The Archaeological Evidence

Boats in the Archipelago before 1500 AD

Significance for the early colonists

A note on galleys

Two Valuable Models

The Berlin

The Breda prahu

The Boat of Botel Tobago

The Whaling Pledangs of Lamalerap

The Lashed-lug Construction Technique

The Contribution of James Hornell

The Broad Perspective in Island

The consequences of the lashed-lug technique

Conclusion

NOTES

REFERENCES

iii LIST OF ILLUSTRATIONS

Figure Page

1. Map of the Eastern Archipelagos with places mentioned ii

2. Basic construction principle 3

3. Balotos and modern 5

4. Boatyard at Tans Beru 8

5. Shell of a with dowels for the next plank 10

6. The method of fitting planks edge to edge 11

7. Pulling the hull together with rotan · 13

a. Details of the prahu belang 15

9. Internal stucture of the pledang 16

10. structures 19

11. A kora kora of the early 19th century 20

12. Model kora kora at 23

13. Drawings of boats from the Alcina MS 28

14. Excavation at , 34

15. Kora kora of Sulu pirates 38

16. Kora kora model from Sangir 43

17. Model prahu from Tanimbar 44

18. Construction of the Botel Tobago boat 47

19. The village of Lamalerap 50

20. A pledang in its shelter 52

21. Detail inside a pledang 55

22. Plank patterns of the bows, 4 variants 60

iv FOREWORD

There are few indigenous records about the boats used in the islands off South-East Asia before 'first contact' accounts by post-medieval European explorers; and excavation and the critical study of other evidence has only recently begun to reveal nautical information. A knowledge of precisely how boats were being built and used at about the time of this first contact, before any European influence, is thus of high priority in historical research. If this state can be established it will form a baseline from which other research may extend our knowledge back into the prehistoric period.

In the early pages of this monograph, Professor Adrian Horridge of the Australian National University, Canberra provides us with an annotated translation of the sections on boat construction in a manuscript, Historia de las Islas e Indios de Bisayas, which was compiled in 1668 by Fr. Alcisco Alcina, a Jesuit priest serving in the Philippines. Internal evidence in the text suggests that at the time Fr. Alcina was writing, western influences on indigenous boatbuilding had not been great. From the 18th century, however, these influences became predominant and examples of traditional Philippines boats can now no longer be found. Fortunately, in the outlying islands of , modern examples are still to be found of what is evidently an old tradition; one in which several details of construction correspond with those described by Fr. Alcina. An examination of these 20th century boats, with details of two 19th century boat models in Berlin and Breda, in the latter part of this Monograph, illuminate some of the obscure points in the Alcina manuscript, and enable Adrian Horridge to establish with reasonable certainty the methods of boatbuilding . used in 17th century Philippines. An earlier and wider use in the western Pacific of these "lashed-lug" techniques seems probable, as Horridge postulates.

Professor Horridge also notes that the few known examples of prehistoric plank boats from northern and western Europe - the Ferriby boats (Wright, 1976); the Brigg 'raft' (McGrail, 1981); and the Hjtlrtspring boat (Greenhill, 1976, 81, 119, 121) - also had planking with lugs or cleats. However, any suggestion that this technique was diffused from a common source to N. W. Europe and S.E. Asia before the European Bronze Age would be highly speculative at this stage of research.

v Adrian Herridge is uniquely qualified to write this Monograph, not only because of his academic and professional experience, but also from knowledge gained during recent fieldwork in the area. Further information on modern boats from this region may be found in Maritime Monographs 38, 39 and 40, and in Herridge's (1981) recent publication on the Indonesian indigenous boats or prahus.

Sean McGrail • Chief Archaeologist

References:

Greenhill, B., 1976, Archaeology of the Boat, A & C Black, London. Herridge, G.A., 1981, The Prahu, OUP, Kuala Lumpur. McGrail, S., 1981, The Brigg 'raft' and her prehistoric environment, BAR, Oxford. Wright, E.V., 1976, The North Feniby boats, NMM Monograph No 23, Greenwich.

vi Acknowledgements

The existence of the Alcina MS was drawn to' my attention by W.H. Scott of St Andrews Theological Seminary, Manila. To him I am greatly indebted for numerous details of the period and for his assistance with the translation. I am also grateful to Fr. Herman Mueller, S.V.D., Rector of the University of San Carlos, Cebu City, for later sending me a copy of the translation by Paul S. Lietz, Loyola University, made in 1962 as part of the Philippine Studies Program (since discontinued) of the University of Chicago. For the models, I am obliged to Mr. Nauta, Conservator, Rijksmuseum Volkenkunde Juatinius van Nassau, Breda, and to Dr. Gerd and Dr. Gerd Hoepfner, Museum f()r Volkerkunde, Dahlem, Berlin. For hospitality at Lamalerap I am grateful to Fr. Dupont, but above all I am obliged to many unnamed friends for their enthusiasm and for real assistance in difficult places along the way. The drawings were made by Chris Snoek and Irma Appleman. Numerous drafts have been typed by Tess Falconer.

The penultimate draft was read by Drs. Brian Fagan, Ian Glover, Campbell MacKnight and James Urry; to these I am grateful for many fruitful discussions. If any of my errors remain, I would be grateful to learn about them from travellers and historians.

vii Introduction

To trace the history of boat construction in islands off Southeast Asia is a relatively easy task as far back as the middle of the nineteenth century. Before that, the records are scattered, they often lack sufficient detail, and what little there is refers to several different traditions that are not to be confused. The manuscript of Fr. Alcina is one major source of information that has been neglected because not elucidated. The details given by Alcina can be directly related to five other lines of evidence, all of which can now be fitted together. These lines are, 1) other early accounts, 2) two archaeological finds in the Philippines, 3) scattered surviving boat technology from isolated places, 4) a small number of old models, each unique, and 5) the principles which ensure a strong construction. Some of this evidence has already been presented in monographs 38 and 40 of this series. I will now present new evidence along all these lines together with the Alcina MS which illuminates the whole process of early eastern boatbuilding.

The conclusion is that one particular construction system was very widespread; it can be traced back to the beginnings of the records and it was particularly well-adapted to the uses required and to the materials and tools available. The design is here called the lashed-lug planked boat: its crucial components are (a) the shell-first construction on a keel or dug-out foundation (b) edge-dowelled planking of hardwood carved to shape (c) lugs carved in situ in transverse rows across the inside of the boat (d) flexible frames placed in tension to compress the planks together, (e) many transverse thwarts also lashed down to the lugs and rib-ends to squeeze the hull. I argue that the dowelling technique required an iron tool and replaced sewing which did not. We can infer that a variety of sturdy seagoing boats of several kinds have been built with this standard construction system over at least a thousand years and over a wide area ranging from southern Taiwan to Timor, with the main centre of ·development in early historical times the Visayan Islands and the Moluccas. An interesting question is how many of these features were known to the ancestors of the people who colonized the Pacific.

Possibly also we are looking at a construction technique that originated in an early South Asian culture and spread in the opposite direction more than 2500 years ago, from which the earliest Scandinavian boats were derived. To infer so much means that the several elements of the technique must have spread in the Indo-European Bronze or Iron Ages, and examples may turn up from Asiatic lake deposits. For other details of these

1 conjectures see Hornell (1936b) and Wright (1976) and the final sections of this paper, bearing in mind that the astonishing conservativism of these boatbuilders inclines those working on the subject to think in terms of diffusion of techniques.

The Method of Construction

The essential feature of the method of construction was that when the planks were carved to shape from the solid hardwood, large projecting lugs were left at regular intervals on the inside surface. When the planks were assembled to form a complete shell, the lugs were shaped and bored with one or two holes. Then, to hold the boat together, transverse thwarts locked against the lugs were pulled together with rotan (figure 2a). Alternatively a flexible branch was lashed first to the lugs at the sides then pulled down with lashings to the lugs at the bottom, also compressing the edges of the planks together (figure 2b). In larger boats the methods in figure 2a and 2b were combined, by lashing to the lugs a rib that pulls the bottom part of the hull upwards to several thwarts (figure 2c, d). The resulting stresses on the hull press the edges of the planks together. Not only does this close the seams and prevent them working as the boat bends and twists, but it is also a clever device which ensures that the compression forces between the planks must be overcome before the seams can begin to open at alL This construction system, or parts of it, can be used for a but in those described here the planks are prevented from sliding relative to each other by hidden dowels at regular intervals within the thickness of the planks. These dowels are subjected to huge shearing loads by the torsional and bending stresses on the boat. The loads are lessened when the planks are long, perfectly shaped to fit together, and then cramped firmly together by the lashings. The dowels alone are insufficient to hold the planks together except in small boats used on placid water.

The three-dimensional interlocking system means that each plank has a predetermined shape which cannot be changed without altering all the others. Also the complex pattern of lugs in rows across the planks is carved in the right place without plans or written records. Partly in consequence, all the boats turned out by one group of men, or one village, are identical in proportions and plank pattern. They need not be the same size because the builders have an orally transmitted system whereby the plank lengths and lug positions are all fixed as a proportion of the length of the keel plus its extensions. Fixing the sizes of all the parts as fractions or multiples of one component is still the method employed in building outrigger and boats over the whole region.

2 (a)

(b) (d)

· flexible rotan-- /

Figure 2

Two components used in clamping together the edges of the planks so that an extended or a planked boat on a keel becomes an arch in compression. (a) by lashings in tension between thwarts that are locked against lugs, (b) by a flexible rib that is pulled down in tension. The arrows show the directions of the forces. These two methods can be found separately and in combination, also in a variety of modifications in many types of canoes from the Indian and Pacific Oceans (Haddon 1936, Hornell 1936), (c) section through a kora kora at Kupang, Timor from Freycinet (1828, plate 24), (d) cross-section of the model at Breda, which is almost identical to the pledangs of Lamalerap, see Figures 9, 21. (c) and (d) show an efficient use of materials by combination of (a) and (b), with the additional device of a lashing between the rib ends and the upper thwart.

3 The Alcina Manuscript

The "Historia de las Islas e Indios de Bisayas" was originally a manuscript dated 1668 by the hand of Fr. Alcisco Alcina, S.J., who served from 1632 to. 1672 in the islands of Leyte and Samar. These islands are in the eastern Visayan Islands which lie between

Luzon in the north and in the south. An early Spanish se~tlement there is now Cebu City. An 18th century copy of parts of the MS (known as the Mu~oz copy) is in the Biblioteca Real, Madrid. The modern Balthazar transcription of this was made 20 years ago for the Philippines Study Program of the University of Chicago and a Xerox copy of that I obtained from the Rizal Library of the Ateneo de Manila. A preliminary translation into English by Paul Lietz of Loyola University has circulated among scholars, eg quoted by Spoehr (1980), but the sections on boat construction are incomprehensible without reference to real boats and models.

Father Alcina provides us with the most detailed of the earliest accounts of the construction of the planked boats in S.E. Asia. There, highly sophisticated boatbuilding techniques were established long before the arrival of the first Westerners, and the continual maintenance of the colonists' was made possible because they found ready-made a industry, which they were able to employ immediately and eventually take over. The were made to fell and haul timber, carve it to unfamiliar shapes and make rope and , but their own shipbuilding and long-distance were suppressed. In fact their splendid maritime traditions were destroyed and replaced by a western one. The building and refit of the Spanish ships in the Islands was essential for the long voyage to Mexico which was the lifeline of the Spanish outpost in the Philippines (Schurz 1939). To the South, an exactly comparable relationship was established from the start by the Portuguese in the Spice Islands with the maritime sultanates of the Moluccas at Ternate, Tidore, Ambon and Banda.

While the contents of the Alcina MS are not new to specialists, his description of the traditional boatbuilding is here presented for the first time in an accessible form in English, together with an attempt to explain every detail by reference to other material. It is not a full translation because the original is wordy and often discursive, but hopefully none of the significant points have been omitted.

A typescript copy of the Balthazar transcript in Spanish has been deposited in the library of the National Maritime Museum at Greenwich.

4 The Manuscript

On the construction, capacity and variety of native boats

Chapter 8

"Although accomplished in all the Arts, the Visayan Indians of the Philippines excel in carpentry". In his first paragraph Alcina praises their ability to build houses, churches and other objects of use and art from wood. He then anticipates his next three chapters on the construction and the variety of boats which they had used from ancient times in their ordinary voyages and in their mangayao raids, as they called them. Mangayao raiding, especially for slaves, was a way of life common to much of island south-east Asia. In the mountains of Mindanao and northern it means "head.hunting". Mengano simply means "attack" in Malay. The largest boats of the Visayans and the majority of boats of all sizes, were primarily plunderers and slavers. In the peaceful business 9f and trade among the islands they mostly used a with a double outrigger (see figure 3), called a baloto by the Visayan natives, and a by the Tagalog speakers (see note 30). In Alcina's time, these came in different sizes, from large ones which carried many people to single-handers.

(a) (c)

j~ I

(b) (d)

Figure 3

(a) and (b), Two balotos from illustrations in the Alcina MS, (c) Outrigger canoe of Jolo, Sulu Islands; from Hornell (1920) plate IV, (d) Coasting vessel with outriggers from Misamis Province, Philippines; from Hornell (1920), Plate V.

5 ''Many times I have watched an Indian emerge from his house, carry his boat down to the water on his shoulders and off, accompanied by his dogs, and he reverses the actions on his return". Alcina recalls the local style of sailing which was watched with admiration by the priests and other Spaniards. "They are quite accustomed to pack 10- 12 people in a small baloto until there is a finger-width of freeboard, and then sail off unconcernedly. The same boat would soon come to grief with only·a couple of Spaniards in it. Evidently the Spaniards are psychologically different from the Indians, because they are certainly not physically different". Alcina says the same about their tiny houses, which he often had cause to visit when on the road, either when travelling with the Sacraments, or to stay the night. On such occasions he was hardly able to breathe; he lost his sense of time and would have to go outside to repeat to himself the words of the psalm "Os meum aperui et attraxi spiritum".

"A balato (canoe) the size of a tree trunk is made as follows. First they fell the tree so that it falls in a convenient way, because afterwards it can be turned over only with difficulty. Then the outside is peeled off to a depth of about two fingers, which is easily done when the tree is green. They then stretch a thread (called kutur) along the tree to serve as a centre line to preserve the symmetry during the work. They remove the upper half cylinder of the trunk as it lies on the ground except for two upwardly projecting pieces at either end which will serve as poop and prow. Using these ends as supports they may later raise the sides of the canoe with boards. They cut the outside roughly to canoe shape by narrowing the ends. Then with adzes they hollow it out, leaving projecting lugs which they call tamboko (see note 1). These will serve as the attachments for the transv~rse thwarts, called agar, which considerably strengthen the canoe and are like a ladder from one end of the boat to the other. The thwarts must be very strong because the other timbers are lashed to them, especially when they build up the canoe with planks, called timbao, to form a larger boat which is then called a tinimbao.

"They then hollow out all the inside of the dugout, taking care to get the sides the appropriate thickness for the length, by drilling holes with a tool called a lokob (see note 2). This is not round like our auger, with a twisted point, but is slightly hollowed on one side and rounded on the other, like a gouge. These drills are made in different sizes, some broader than a finger. They are about a palm long with a handle of hardwood. Two pieces of wood are thrust through the handle in the form of a cross so that it can be turned while it is struck with a mallet (pakang). As they twist the gouge through the wood its iron tip cuts out a clean round hole. In the bottom and sides of the dugout one sees many of these test holes to reveal the thickness that remains as they 6 chop. They have a square adze like ours (daldag) and a skewed adze called a binkong (see note 3) which are both essential for thinning down the bottom and sides of the dugout to the correct thickness. They work away, mainly with the axe, adze and drill until the dugout is perfect. Alone, a hardworking Indian can finish a large canoe 5-6 arm spans (8 to 9m) long and 6 or more hand spans wide in 8-10 days. Usually however, they take months on account of their unbelievable laziness and insufferable slowness".

Alcina then turns his attention to the larger boats which are built of planks. Of these there are a variety of types, each in a range of sizes. He explains, establishing his credentials, that he has arranged for the building of more than 20 large ships and he has observed those built in the native style. Alcina was a superintendent of boatyards and he explains that missionaries must often develop a skill, for otherwise "those who don't know how to do such things don't have anything".

To make a planked boat, the keel and the two posts are selected from an especially hard timber called tugas. The required size of the is given to the men who go off to the mountains to cut down the trees. They take only two planks from each tree, being partly restricted by lack of tools to do otherwise (see note 4). They reject trees which have a rotten core, called bokag, which is useless for planks. Also the outer layer of the trunk, the sapwood, called aramay, is useless as it soon rots and breaks up, especially (he says) if cut in the last quarter of the moon, after the 22nd day or before the 3rd day (during the amero). The planks are cut by moonlight from the solid trunk by a man who knows the correct twist, called lubag. There must be extra twist in the plank next the keel, called the dokot, and in the second from the keel called the lonor (see note 5). These garboard strakes determine the shape and performance of the boat, for if the bottom is flat it will have greater capacity but be slower. When the hull is very narrow the boat is not readily steered.

"In the forest they cut all the planks to the right length and never allow short planks to be extended by another piece as is done in Europe (see note 6). For small boats they lay a square keel, shaped like a dugout canoe (see note 7). For a boat longer than 10 arm spans (17-19m) a keel like a dugout is not satisfactory unless it is very thick and flat. They mortice (see note 8) the stem and stern posts flush (a sa ni vel) on the ends of the keel, so increasing the length of the boat by half an arms span at each end (see figure 4). They incline the stem and stern posts in this way so that the planks join with them to form a strong design, (see note 9). They assemble the planks, steadily improving their shape, first with an axe and then with an adze. On each plank they leave the projecting

7 Figure 4

A boatyard at Tana Beru, South Sulawesi 1978, showing keels already to be built into shells. Curved grown timbers for ribs, etc., are waiting to be selected. Except that some of the wood has been sawn there is no sign of western influence in this picture.

8 lugs, called tamboko, as in canoe, about the size of a man's hand, separated by a maximum of 5-6 palms from each other, projecting by three fingers' width in the upper part and flush with the plank at the bottom (see Figs. 2, 9 14). Besides adding some strength to the plank, these lugs act as the seatings for the transverse thwarts or for rounded branches which run (across the boat) from plank to plank. For this purpose, they have two holes through them for the rotan lashings. The thwarts are lashed to the lugs exactly as in the dugout canoes. In the transverse thwarts, lies all the lateral strength of the boat, which is held together entirely with wooden dowels and rotan lashings, with no iron nails (see note 10).

"To hold the planks closely to the keel and also to each other so that they stay together without further bracing, they have an ingenious technique that is not known in Europe but is certainly effective for their purpose. With their drill, the lokob (see note 2), they make round holes in each side of the keel and in the edges of the planks about one handspan apart, from stem to stern of the boat. In every hole they fit a dowel the size of a finger, of a tough, hard and rotproof wood called bahe. In places where there is no wood as good as mangrove they make them of red Brazilwood (sappan, see note 11). The dowels are fitted tightly into the holes so that they look like a comb, or row of tusks, each projecting about one hand or less (figure 5). In the opposite edge is the corresponding row of holes into which they fit as the planks are brought together edge to edge (see note 12). Afterwards we will see that the planks do not stir or come apart under the strongest buffetting of the sea. Planks are added by this technique until they have the hull the right height. They have two types of , depending on whether the boats are high or low".

"Having assembled the boat with 5, 6 or more planks, they leave it out in the open to season for a month or two, because continuing with the wood still green would cause warping to develop later and spoil the work. When thoroughly dried out they dismantle it plank by plank down to the keel, renew the broken dowels, of which there are usually plenty, and reshape the garboard strake so that it fits closely on the keel without being forced. The process of refitting the other planks, called sugi, is done as follows (see figure 6).

"A small piece of wood is cut about the size of the palm of a man's hand. On one side of this block, there projects a little lip which fits between the two planks, resting against the lower one. On the same side, in a position exactly related to the tongue, a small spike of iron projects as much as the thickness of a large coin, and is firmly held in so that it cannot shift. A man with a strong grip grasps the wood, with the tongue against the lower edge. The iron point is on the upper plank".

9 Figure 5

An early stage in the growth of the shell without lugs, as in most prahu building now in Indonesia. The planks are held together edge-to-edge by large dowels of mangrove wood or preferably sappan wood (Caesalpinia sappan) spaced about 20 em. Stiff ribs will later be carved to shape and fixed in place with wooden pegs (treenails).

10 (d) :-·--.... (e) (a) I I I I I I . ·~-c .. -···I ··.JI ' '-· .....: ..... -~~ (_·-~~

Figure 6

The sugi process, showing how the contour of the upper plank being added is matched to that of the lower one already there when the relative positions are fixed by the dowels. It is done by making a negative image of the lower edge upon the temporarily fitted upper edge, by making a scratch on the upper plank with a tool that runs along the lower one. (a) The tool is described by Alcina as a block of wood with a projecting tongue located at a set distance from an iron spike which projects a little. The tool illustrated is a drawing of that (called a kilik or kirik) used nowadays for exactly the same purpose on the north coast of Madura, Indonesia. (b) The tongue is run along the upper edge of the lower plank and a line parallel to this edge is scored on the upper plank. The rubbing surface of the tool is the back of the tongue, marked by an arrow. (c) All below the score mark is then cut away with a chisel, leaving a perfect fit. (d) The edge is hollowed to take the fibrous stuffing. (e) Modern steel tool (singkolo) of the same function, now used in many Bugis boatyards. The boatbuilder has a set of different sizes threaded on a string.

11 The MS, obscure at this point, reads as follows:"corriendo con fuerza en la parte superior o tabla que se encaja una raya o seP:al, la cual van con un escoplo siguiendo y ~itando lo que queda debajo de dicha ray par el un lado y otro de la tabla dejandole alga acanalado el bordo: porque en aquella canal o vacio encajen el baroc". I translate "running the tool forcefully along the upper plank he scores a line. The scoremarks are then gone over with a chisel which trims off the wood below the line on both sides of the plank leaving the edge somewhat hollowed out, where it is later filled with the palm fibre". We must infer that the object is to achieve an exact fit. I interpret the above to be a description of a tool which runs along one plank and scores the other plank to transfer the contour of the lower edge exactly on to the surface of the upper one. The fit has to be achieved in this way because the dowels have already been located (Figs. 5 and 6). The similar tool used in Indonesia today is the singkolo (Fig. 6e) and in Madura it is the kirik (Fig. 6a). The hollowed-out channel might be anywhere on the face or the edge or it may consist of several grooves on the surface where the two planks meet.

Continuing: "In this channel they fit the barok, a soft fibre-like goats' hair, from a palm tree, as a filling for the edges, the mortice holes and the tongues of the planks which fit into the stem and stern posts (see note 13). Like oakum the barok makes the seams waterproof when it is soaked in water. The planks are replaced one by one with the same careful use of the sugi technique until the hull is a complete shell for the second time. For a large boat the process is repeated.

"To adjust the planks more exactly they use a technique they call osos. Rough logs from the forest are placed across the gunwales and strong rotan is taken from one end of each log under the keel and up to the other end of the log (Fig. 7) where it is bound tightly. Then they hammer wedges into the ends of the logs, splitting them and making them thicker so that the force is transferred into a compression of the hull, and the edges of all the planks are forced tightly together until their edges fit perfectly and the whole boat appears to be made in one piece. While the boat is held together so, they drive little wooden nails (see below, ipil) transversely through the dowels, both above and below the joint between the planks. The care with which they fit the planks strengthens the shell because it is all knitted into one piece. They call the process of plugging up the joints, pamota, and when closing a large crack they say "napiungna" which means shutting the eyes. Metaphorically the seams are closed together like eyelids." This technique, it should be noted, is essential for a boat constructed as described, because it cannot be caulked. Cracks in modern boats with thick ribs are commonly filled by forcing in packing such as old sail cloth.

12 Figure 7

This photograph illustrates the osos technique, which depends on the force exerted by the wedge in the split log. Every other detail in this picture shows western influences. This hull, under construction on the island of Bangka, contrasts with the traditional hull in the long straight keel set at an angle to the stem, the alternating ribs and floors, and sawn planks nailed into a broad apron on the stem. Here the thin planks are bent around ribs and floors that are previously set up. The hull shape and these features show that recent Malaysian boatbuilding techniques have spread into Indonesia as far as Bangka.

13 "To complete the shell they drill holes with a small iron spike which is square like a ke y, with grooves in it, and a wooden handle as described for the lokob. This drill goes in easily like our augers. They bore straight through each dowel where it is outlined on the outside by a score mark (see note 14). In these holes they insert splinters of a hard wood called ipil, which seems to be, always green, and which in water produces a resin which stays sticky. The splinters, two fingers long, are left as locks through each dowel, so that it cannot shift. Since these locking splinters are squared and sticky they never come out. In fact, the dowels break before the locking pins. In this way the shell is finished without iron fastenings".

Chapter 9

How they continue, up to the launch

"The way they complete their boats also does not coincide at all with the European method. Along every plank from stem to stern, we have noted the lugs or tambokos which are spaced every 5-6 hand spans. These are placed in such a way that across the planks they correspond in number and size from one side of the boat to the other. When finished they look like little staircases that run up the inside of the boat. Two holes are made in each of these lugs, without passing through the side of the boat. A round branch is lashed to each staircase, running across the planks from binding to binding. Depending on the size of the boat, this branch is never thicker than the upper arm nor thinner than the wrist (Figs. 8 and 9). On the other hand, they take care to use pliant, durable and light wood of which there are many kinds (see note 15). When each piece of this wood is scraped and well-rounded they make a hole (or two holes for a large one) at each place where it crosses between the tambokos (see note 16). Special braided rotan called talolora (maybe tali lora) is threaded through these holes, binding the round branch to the lugs so that it cannot be pulled one way or the other in the lashings. This construction holds the planks together from stem to stern (see Figures 8, 9 and 17). On all the planks they attach the thwarts, called agar, between the two sides of the boat, and each of the thwarts is lashed to the next at three or five different distances from the keel, until all the boat is bound together (see Fig. Zc and d). Although the boat is not strengthened by our methods, it is just as strong and in fact better for this part of the world. Experience shows that on account of the high humidity and of the warmth of the sea, iron here rusts so quickly that nails become thin and rot the wood. Day by day (on our own boats) the planks rot where nail holes open up and let the water in, but that never happens with their boats".

14 Figure 8

Drawing of part of the inside of a prahu belang of the Aru Islands, e xamined in 1977. This boat is constructed with flexible ribs lashed by rotan to thwarts and internal lugs as described by Alcina. The boat is more fully described in Herridge 1978 and 1981.

15 Figure 9

A plect.lg of the whaling village of Lamalerap, Lomblen, illustrating in almost perfect form the details of the lower hull construction described by Alcina. The tension in the lashings of the rib is carried through the main thwart and through the lugs on the plank to convert the hull into an arch in compression. This pledang is partially dismantled; in use it has a platform extending over the bows supported by long (Fig. 20) that are lashed over the forward thwarts. The boat is carried down the beach by the ends of the thwarts, which are convenient handles.

16 Alcina then explains at length that other authors are wrong when they suggest that iron fastenings are not used in that part of the world on account of magnetic rocks which would draw the iron out of the boats. He mentions that iron is sometimes used for attaching a in the western fashion, and that they store their vessels on large chocks on land to prevent rot and attacks by white ants. He then continues with the description of the outriggers and the structures which support them. Again the MS is obscure so that an exact interpretation is difficult from it alone, but we can draw on other evidence to infer what Alcina is talking about.

The Spanish reads as follows:

"Amarradas ya las tablas todas al modo dicho con sus agares, atraviesan cuatro a cinco a mas llavetas a cuadradas a redondas de bordo a bordo en estas Embarcaciones

dejandoles fuera de cada lado 0 bordo una braza 0 mas de largo a dichas llavetas y en ellas, habiendolas primero amarrado muy recienmente con todos los palos atravesados o Agares que estern dentro del Buque del Nav(o y sirven como de bancos, aunque el mayor y ma's fuerte es esta Haveta, que eHos Haman Batangan, que en esta lengua quiere decir: cosa que est{ en fiel y drecha, ponen ciertos palos de mangle par ser madera recia dos comunmente en cada batangan, a Haveta asientan en los lados de eHas arrimados o cabezeados con un tarugo cuadrado y recio que atraviesa las puntas de dichas batangan en que amarran reciemente los dichos palos que Haman Tadic: estos son comunmente tuertos hacia la parte baja como una ese, y en ellos, que ya con la comba llegan cerca del agua, ponen a lo largo dos o tres, y a veces cuatro canasN de la grandes de aca/ a un lado y otro, de modo que quedan ladeadas y hacen un modo de ca'Xizo, que vendr/ a estar unas dos brazes o cerca de ellas par cada lado lejos del cuerpo o buque de la embarcacidn, cuanto mayores apartan mas estos que Haman: Cates, que sirven de llevar contrapesada la embarcaci6n que aunque ladeen es cora en ellos y sirven de sosten para los balances; porque con estos cates, que son como barandas a un lado y otro, y a la lumbre del agua son mucho menos par recibir las ca~as primero, y con eso se ladeen poco la embarcacion, prevencion necesaria, y mcfs en los mares y alas grandes que hay en algunas partes de estas Islas, de que ya dije alga arriba, que sin este reparo o estribos se zozobran no pocas cuando no esta'n bien contrapesadas y fuertemente atados dichos Cates".

A literal translation is as follows:

"After the boards have all been fastened together with their agares in the said manner, they place four, five or more transverse beams either squared or rounded across from 17 side to side in these boats, leaving these beams extending an arm-span or more on each side and after fastening them very securely to the thwarts which are inside the hull of the boat and serve as benches (see Fig. 9), although the largest and strongest is the beam which they call batangan (see note 17), which in their tongue means a thing which is always dependable, they place some poles (sticks? beams? etc) of mangrove wood on them - because it's very tough - usually two to each batangan or llaveta. On the sides of them, upright or tilted, they seat these poles, which they call tadik, with a tough square peg which goes through the ends of the said batangan to which they fasten them tightly: these are commonly twisted downward like an S, and on them, reaching almost to the water as they do because of their curvature, then put two or three, or at times four, bamboos of the large size found here, lengthwise on both sides, in such a way that they are tilted and form a kind of hurdle along each, side which will be some two armspans away from the body or hull of the boat - the larger, the farther apart - these which they call kate, which serve to maintain the balance of the boat which is supported upright on them even though they dip down and they serve to counteract

rolling bec~use with these kate, which are like balustrades on both sides and at water level, the rolling is much less since the bamboos receive (the force of the wave) first, and with this the boat heels over very little, an especially necessary pr ecaution in the and great waves found, in some parts of these islands, of which we have said something above, so that without this protection or support, no few capsize when they are not well counterbalanced or the kate not firmly attached".

Let us now bring in some outside evidence. Father Jacobs (1975) translates the 1544 Portuguese account of Antonio Gal v~o (see note 20) as follows: "When the boat is finished, they set across it from side to side ten or twelve well-wrought cross-beams, which serve as laths or like the bands of a , and are called ngadju. When they have been set athwart and steadied, they stretch out beyond the boat on each side, one, two or three fathoms according to the ship's bulk. And to these ngadjus they fasten parallel to the boat two or three rows of canes, which are called kangalha and on which the oarsmen sit down over the water, apart from the other rowers inside the hull of the boat. And at the very end of these ngadjus are some wooden forks, called pa9-J, to which they tie other thicker and longer canes, which they call samah, to support them when they heel". Boats of this type have been frequently described and sometimes illustrated from the area of the Philippines and particularly from the Moluccas over the past four centuries. Some are described by Horridge (1978, 1981). Other recent examples are reproduced in Figures 10 and 11.

18 (c)

(b)

outrigger

outrigger

Figure 10

Illustrations showing the 5-shaped outrigger connector (a) in a canoe built up with one extra plank from the Sangihe Islands, just south of the Philippines (Steller and Aebersold 1959). (b) attachment of the outrigger float at the island of Talaud, also south of the Philippines (Nooteboom 1932). (c) an (10m long) from the island of Bacan off Halmahera, shown in section with lugs on the planks (Hornell 1920). Canoes which face rough conditions at sea have a strong connector which is able to adjust the height of the outrigger float below the outrigger . Direct attachment (Fig. ·3c) e.g. at Nias Island, and nowadays at Menado and many Philippine Islands, is less effective. The design of the connector piece is specific for every region (Nooteboom 1932, p. 217).

19 (a} ;\ I-' - I r _ , , - -"r ·-- 1_..- r": .... I 1. - I • .\ II I... v~ r . ~ ~ ~ A ~--- v \ I r-

Figure 11

The kora kora. (a) The usual design seen from above. (b) The apparent arrangement in (c) which was possibly under repair. (c) A kora kora ashore at Manado, North Sulawesi (Celebes) in 1828, redrawn from an illustration by Paris (1843). This vessel was probably being repaired because, as in (b), the platforms on the outriggers do not run the length of the boat. The tripod mast has been removed and the absence of outrigger floats is quite customary when beaching. Possibly oars were poked through the holes in the long side boards. Note the plank pattern, the lack of a distinct stem post and the 12 transverse outrigger beams (batangan) in two groups, from Paris 1882-92 plate 322.

20 Before AD 1500 plenty of reasonably large boats were built without outrigger floats. For example, 8th century Javanese boats illustrated at Borobodur, 9th century boats of Austronesian speaking Chams at Angkor Vat, Cambodia, and as mentioned in Chinese accounts; references in Horridge, 1978.

Later on, many such are illustrated, as in Figures 2(c), 15 and 16.

As the local people progressively copied the stable western hulls, with torsional rigidity and ballast, the use of outriggers waned.

We can continue with a free translation as follows:

"When the planks are all held together by the internal structure, they place four or more squared or rounded beams (batangan, see note 17) across the boat, extending by an arms span on either side. They fasten these down to the thwarts inside the boat. On these transverse beams they attach lengths of mangrove wood called tadik, bent in the shape of an S and fixed with a strong peg to the end of the transverse beam (see note 18). The tadik twists down towards the water, not necessarily in a vertical plane, and down there they fix several large bamboos as outrigger floats parallel to the hull and about 2 metres from it. The parallel group of bamboos form a sloping array (possibly arranged along the sloping outmost curve of the S, see figure 10). The outrigger floats (lcate, see note 19) balance the boat as they dip into the water."

The following extract from Dampier's account of the Philippines, suggests that the outriggers touch the water only when the boat rolls.

"The proes that are built for this purpose, are large enough to entertain fifty or sixty persons or more. The hull is neatly built, with a round head and stern, and over the hull there is a small slight house built with bamboes; ••• Besides this, they have outlayers such as those I described at ; only the boats and outlayers here are larger. These boats are more round, like a half-moon almost; and the bamboes or outlayers that reach from the boat are also crooked. Besides, the boat is not flat on one side here, as at Guam; but has a belly and outlayers on each side: and whereas at Guam there is a little boat fastened to the outlayers, that lies in the water; the beams or bamboes here are fastened transversewise to the outlayers on each side, and touch not the water like boats, but are one, three or four feet above the water and serve for the barge-men to sit and row and on: the inside of the vessel except only just above and abaft, being taken up with the apartments for the passengers. There run across the outlayers 21 I

two tier of beams for the paddlers to sit on, on each side of the vessel. The lower tier of these beams is not above a foot from the water, so that upon any the least reeling of the vessel, the beams are dipped into the water, and the men that sit are wet up to their waist; their feet seldom escaping the water. And thus, as all our vessels are rowed from within, these are paddled from without". (from p. 15, 229-230 of the Argonaut Press edition). Dampier may have described the daramba which was a catwalk over the water for an extra row of paddlers, and possibly he did not refer to the actual floats.

Alcina mentions that in his time double outrigger boats were known to the Spanish only around the Philippines, but single outriggers were found in islands passed by the on their long voyage from Acapulco, probably the . He praises the use of rotan, which is abundant and easily replaced at sea. He continues:

"On top of the cross ties (Haves, an old Spanish boatbuilding term) on which are stretched the supports of the outrigger booms (kate) they put three, and on large boats five, walkways (erradora), which improve the strength and facilities on board (the lateral walkways are called pagquirai, see below). The walkways also protect the crew, because it is their custom to paddle into battle against their numerous enemies. The central walkway from stem to stern along the middle of the boat is called the btrutlan (see Figures 12 and 16). It does not go all the way in smaller boats. The burutlan has a plank on edge like a railing along each side, with a gap beneath where the rowers (below) put through their paddles. Its floor is a plaited split cane mat which is sufficiently strong to walk on but much lighter than if made of planks. The men who manage the sails, the passengers and cargo that must be accessible, travel on this central gangway. Battles are sometimes fought up there, as well as on the ether catwalks at the sides. They make more space there in times of war. Where there are men below at the paddles, the floor is raised so that they can stand up, enter and leave, and store the cargo. They call these catwalks the pagquirai which compares them to the eyebrows of the boat. The two lateral catwalks are narower than the one down the centre, but are still adequate for a man to sleep upon. They have railings decorated with regular carvings so that they are a delight to see". By searching through old pictures (eg Figs. 11, 13, 15) and old models (Figs. 12, 16, 17) we can find examples that illustrate the details that Alcina describes.

"Another feature, not typical of western boats, is that they cover their boats with an awning of removable plaited kayane of cane, rotan and palm leaves, which provides essential shade and shelter from the rain. Even the smallest boats can look like houses, but the roof and walls can be quickly folded up when the wind is from ahead.

22 ( Figure 12

This old 19th century model from the Sulu area illustrates the tripod mast and the internal structure that Alcina describes. The platform along the centre line of the boat is the burutlan, the two across the ends are the daramba in Alcina's account but otherwise the kora kora had no deck. This model had no outriggers, and is incomplete in that it lacks platforms (paqquirai in Alcina) for men at the paddles.

23 "In addition to the three catwalks along the hull, the largest boats have additional ones, called daramba, of split canes, which run across the ends of the batangan, above the outrigger floats (Fig. 12). Large boats even have one or two walkways in the midst of the supports on the outside of the boat. All of these types carry two banks of paddlers on the inside of the boat and another two banks on the outriggers. The largest of them have six rows of paddlers on each side of the boat. We cannot match this with our western boats because we don't have the outrigger structure for them to sit upon (see note 21). One cannot deny that all this superstructure occupies an enormous space and causes wind resistance, although the paddles are a convenience (when there is no wind). With six banks of paddlers, or at least four on each side, with 10, 15 or 20 men in each bank, boats can total 100-200 men at the paddles."

"All their boats carry one or two masts, with square sails that can be set in different ways. Our own Visayan Indians used to operate, and still do, with square or rounded sails. Usually the mast is a tripod of three bamboo poles which meet at the top where there is a hoist for the sails (see note 39). The two side poles separate at the bottom, where they are pivoted with wooden pins in two tabernacles so that they can be lowered or raised as convenient. Even when the mast is a single bamboo pole, specially selected to be light and strong, it is pivoted in a cradle so that it can be raised or lowered depending on whether the wind is favourable. The mast supports a rig a bit like that of our Mediterranean saetia (a pinque, see note 22) but they rig their boats in the same way as t~e Lutaos (see note 23), like the Joloes of Mindanao and other peoples in these islands.

"The only rig they use is the one they call bo~oson, in which the sail is stretched between two yards (see note 24). The lateral masts of the tripod stand upon the gunwales and the third ending in the bows acts as a shoring stay so that the tripod is very rigid. The sail is wider than it is tall. An upper yard of bamboo supports the sail: the lower one pulls against it with the sail spread between the two. When not in use the sail is rolled up round the yards. This type of sail luffs up into the wind very well and is all the safer for being set rather low". (see note 24).

"A great number of superstitions connected with boatbuilding have not been mentioned (see note 25). Even when cutting the planks they count the knots on every plank. From the number of knots and where they lie in the boat they suppose that they can foretell how manoeuverable and how lucky it will be. There is a special ritual when they launch a boat for war or for their thieving expeditions. All the way down to the water they put rollers, called hanglar, on which the boat will slide, but instead of the last roller they 24 put a prisoner of war. As the ship passes over him he is squashed into a pancake, and the sacrifice ensures that the ship and all who sail in her will be feared by the enemy and valiant in taking captives. The sacrifice is called the bakalag. In olden times human sacrifice was the custom in Ibabao, Quiguan and other parts of Calagan as a petition for the health of some great man who was ill. They believe that the boat will be more formidable by the power of the blood that is shed. What a barbaric Godless lot they are! A Visayan proverb "Duoharluksin'iginbabakalagna' says that the sacrifice spreads fear like that suffered by the man. Anyway, having got i::he boat into the water, let us describe how they manage it with sails and oars".

Chapter 10

On boat types, sizes and handling

"The commonest of the small boats that they used for their ordinary affairs in olden times, and still today, is the baloto, which the American Indians would call a canoe. These are dugouts carved from a single log in a variety of sizes. Some are so small that one man can carry them on his back as described. These can sail on the sea at Ibabao, which has the worst waves and chop among these islands. It is almost incredible to see one of these tiny boats with two men aboard flying over the waves like a bouncing ball (see note 40). Often they are swamped but the water is quickly bailed out with the paddles. If necessary the men jump overboard until they can empty tlie boats, then they quickly jump in again. I have been in one of the larger sizes of these boats in a hurry across the Ibabao sea with a dozen or more men (called bariga) at the paddles, which they call bugsai. As soon as a wave swamps the boat the paddlers all jump out leaving me sitting up to my neck in water, soaked. Quickly they bail the water out, then jump in and continue the journey. The men must be able to swim like fishes to manage a boat in this way, but even so the danger is not negligible.

"There are larger boats, some very capacious, made of a single trunk. They are light and fast with many paddlers, often carrying little cargo. There are some carpenters, pandai as they are called here, who make them so fine and graceful that they go like darts.

"There is a peculiar kind of canoe called a talimbao which can have one or two planks added on each side to make it large enough to carry cargo. This type is commoner in the island of Boo! where, instead of planks they use bark from a tree called opak or daopak. Having stripped the bark from a whole freshly felled trunk, they leave it out in 25 the open, rolled up inside-out to dry. After it is opened out, pieces 2 or 3 palms in width are cut, of a length suitable for the boat. Splinters of cane are then used to join the edges to the side of the canoe. The edge is then strengthened by hemming with split rotan that holds the cane pins in place. The result is a capacious long narrow cargo boat which floats with about half the width of the bark submerged (when loaded). Although the bark skin is no thicker than a coin it is remarkably tough, strong and durable. So well are they sewn that canoes built in this way, with no tar or pitch in the seams, let in no water, and it is remarkable that they never break up in the waves although fixed only with cane splinters".

"Of the other craft, the smallest of all built skilfully of planks like a proper boat is the balasia, which cannot carry cargo but is used as a fast passenger boat without provisions, especially for going to parties• . To carry cargo such as rice and other goods they add the strips of opak bark. This is an easy operation and a few light bark strips add a great deal to the capacity. For an excess cargo they raise the sides even more with mats of plaited palm leaves all round the sides (see note 26). Even a thing so soft as a leaf, when plaited together and added like a plank to the gunwale, is sufficient to resist ordinary waves, though in severe seas they are as inadequate as the strips of opak bark.

"The barangai, so called, is quite a large boat, and the fastest in these islands, built with carved planks on a square keel and always propelled by great numbers of short paddles called bugsai. These paddles are 6-7 palms long, with a blade shaped like a furnace shovel, not rounded but pointed and with a fine shape. The blade of the paddle is about 2 palms long and a palm wide, flared out towards the point and narrower in the lower part of the blade. The paddle is made of one piece of wood, with a round handle the size of an ordinary walking stick, of strong pliant wood that is not easily broken. A crosspiece across the end of the handle is held in one hand as they pull, as is done with shovels for moving wheat over there (in Spain). Each man makes his paddle to his own taste. It is an effective paddle that moves a lot of water and gets the boat along splendidly. They naturally face forwards when using these paddles, whereas with larger oars (see note 27) like those in European boats, they always face backwards to row.

"On large boats, the paddlers high up use long paddles, called gaor, that have flat round blades larger than dinner places fixed on a separate pole of a length that is matched to the height of the boat (see note 28). The paddle is attached flexibly to the side of the boat, not by a thole pin as in ours. They are loosely lashed between a pair of crossed bars which stick up from the edge (of the catwalk). The paddles fitting between these 26 vertical bars cannot jump out, however much force is applied to them. There are never two or more men on one oar; every man pulls on his own. On first seeing it, people are amazed that these Indians can paddle from dawn until dusk, and the next day, and many more, always with the same force, eating moderately and without resting, as if born to the practice. The fishermen especially seem born to this exercise, whereas those from the mountains are soon tired. It is significant that the fishermen are accustomed to this work from their youth. Even the very small boys have paddles made for them by their fathers in proportion to their size.

In the next paragraph Alcina says that the Philippine natives occasionally use oars called gaoing which are exactly like western oars, but that they do not use them well. They prefer their own gadre (see note 27).

"Boats called biroko are designed with high sides for carrying cargo (Fig. 13). The ones they formerly made were of poor carrying capacity but after the Spaniards arrived they made larger ones. In olden times there were not so many Chinese junks, which are so different from our ships. In the birokoa they carried the loads of tribute, rice, cloth, wax and wine (probably fermented palm sap, Ed.). I remember seeing several biroko as large as a Mediterranean sahetia (seatia, see note 22) and able to carry as much cargo. They are assisted by paddles, which our ships don't have. Only a few biroko are in use at present for their own trade because they now have other boats better suited to cargo".

"The karakoa is a traditional boat still commonly used today. It resembles the Spanish , although heavier than them because of all the daramba, kate and txru~... These are the best boats for these islands and the most suitable for a war fleet. Traditionally they were used for the common practice of slaving, and pillage of the neighbouring islands when the people were pagans. The karakoa resembles a small galley (see note 21) but with more men, as it carries 3 to 5 banks of paddlers on each side. At the present time fleets of them are being built in the name of His Majesty and at his expense, with Indian paddlers and Spanish men-at-arms. One sails in these with greater comfort than in other boats, because aside from their drawing less water than the galleys and brigantines in Spain, everything can be roofed over with the kayane which has already been mentioned. It cannot b'e denied that the ones the Spaniards build are today much heavier because they add conveniences which increase the weight and bulk. The karakoa of our Filipino enemies are still made in the old style as before the Spanish appeared. They make a mockery of ours because they are faster and soon escape. In fact, they are rarely beaten at sea because the differences between theirs and ours is like the difference between a hawk and a hen (see note 29)." 27 Rombo, o Correlon

Figure 13

The group of sketches in the copy of the Alcina MS, but not necessarily by him, or contemporary, or even drawn from life. For example, the supports for the outrigger floats are not drawn as described in the text, with a tadik. The karakoa (lower centre) is embellished with the purely Spanish symbols, the cross, the pennants and carvings at either end. The stern rudder, raised stern and single masts of the biroko are apparently copied from the Spanish. A banka tilimbao is a banks (see note 30) built up with planks (timbao).

28 "All these types of traditional boats are in use today. The Visayan Indians excel in the art of boatbuilding, being naturally skilful carpenters (panday) even in the construction of our own galleons. Maintenance of our boats is continually going on to sustain the trade with Mexico, on which the whole of the Spanish East depends. Of the Indians, only the Visayans are hired as carpenters in our boatyards with no further questions asked; others are hired only if they have been official carpenters previously. There are Visayans as competent as Europeans in selecting and working timber for (Spanish) ships, earning more wages than many Spaniards do. A former chief at the town of Palapag could tell of many. One of the Visayan boatbuilders, Don Juan Polakay, who died a few years ago at a great age, once told me that he has assisted with more than 20 large galleons, not to mention small ones. He had been paid as a chief clerk. Where I was engaged in finding Indian labour the first thing I was told was that I should consult Don Juan Polakay and another Indian called Figuman because of their experience in fabricating any of the goods required in the East.

"I must mention the lightness and speed of the traditional boats of the Visayans. A merchant of these islands, Pedro Mendez, had an extremely light barangar, which went so fast that no-one can stand up in it with two rows of paddlers at work (one on each side). They could go from Pueblo de Paranas where these barangar were built, to Cebu City, a distance of 40 leagues, by from dawn to dusk. This is more than 4 leagues per hour on average (see note 31). It is my own experience that a man running on the shore cannot keep up with these boats".

Alcina was obviously impressed and concludes that they must be able to do about 6 leagues an hour at maximum speed. He makes a final comment about the rudder design before concluding the chapter, saying that in his day they had already adopted the western style of central stern rudder (see note 32), but mounted without iron. The single terminal rudder was slung on two or three strands of rotan, which is more convenient than pintle and gudgeon, because it can readily be replaced at sea. He mentions that on the islands of Mindanao and Jolo, local boats called jl&lgB (see note 33) have two , some even have four. Because in his day this traditional style was found among the untutored Indians (note the spread of technology transfer), Alcina described these rudders in the section on the enemies of the Spaniards. Unfortunately that section of his MS has been lost.

29 Other early accounts

Considering that the early western explorers were sailors and must frequently have encountered local shipping, satisfactory accounts of traditional boat construction are very few before the nineteenth century. The explorers were interested primarily in items for trade, secondly in local customs and politics, and only incidentally in the local crafts (see notes 36 and 37).

An account of the construction of the traditional boat of the Moluccas, the kora kora was made in a MS of about 1544 entitled "Historia das Molucas", which is probably by Antonio Gal vao, and is translated into English by H. Th. Th.M. Jacobs, S.J. The relevant section has already been published in this series, along with an English account of the kora kora from the first Dutch expedition (see note 34). In the same monograph less ancient accounts of Moluccan ships of similar types were also surveyed (Horridge, 1978).

Many of the techniques that Alcina describes can be seen today in the isolated boatyards of the Bonta Bahari of South Sulawesi described in a previous Monograph (Horridge, 1979). The continuity of traditions may be sensed from figures 4, 5 and 6. We can well believe that the Philippine boatyards would not be very different in Alcina's time.

I recently came across two other nineteenth century accounts with similar details to those given by Wallace (1869). In his report of his voyage, Admiral de Freycinet (1828, plate 37) illustrates a kora kora with the aristocrats lazing on the upper deck (burutlan) and internal details of a boat under construction at Timor (Freycinet, plates 23 and 24). The projecting lugs are visible on the inside of the shell; a section of the construction is reproduced by my Figure 2c.

The other account is by the Governor of Ambon in an effort to promote whaling, which in Indonesia had been in the hands of the Americans and Australians since about 1820 (Mussenbroek, 1877). After describing the construction methods and discussing the merits of the lashed-lug boats of the Moluccas which were like those described here, Mussenbroek concludes that western-type cutters would be more suitable and he mentions that at that time a with a western hull had been made for a Mr. Beth in Surabaya. This shows how late and limited was Dutch influence and it sets a maximum age on the , a western or cutter now common in eastern Indonesia and still

30 built as a shell of planks fixed together by internal round dowels, with western type ribs and floors inserted later (Horridge, 1981).

There may be other accounts to which I have no access, in archives in Dutch, Spanish or Portuguese, or in Chinese, old Vietnamese or from the Indian trade ports in Gujerati. In fact very little data on prehistoric boats is available from Asia at all. In my experience, the old accounts such as we have are explicable only by reference to real boats, but now at least it will be apparent that great care in translation and interpretation of words is required. I provide a basic description of the lashed-lug type of hull construction which was widespread in the sixteenth century. The evidence suggests that boatbuilding by something like the technique described here goes back far beyond that and was in fact a prehistoric skill of the Austronesian speaking peoples, although before the arrival of metal tools the planks would be sewn rather than fixed by dowels.

31 The Archaeological Evidence

Four planks of an ancient boat, discovered by illegal porcelain hunters, were excavated in October 1978 by P.S. Gonzales the city engineer of Butuan, Philippines, and subsequently taken over by Dr. Jesus Peralte, Curator of the Anthropology Division of the National Museum of the Philippines. Nine planks from another vessel nearby were systematically salvaged in 1977 by Cecilia Salcedo of the National Museum and eventually carbon-14 dated at the University of Tokyo. The date of the second boat was 1200-1300 AD and the first boat lay more than a metre below a layer of refuse containing Sung porcelain. The boats were about 15m long (Fig. 14). The data is from the Anthropology Division of the National Museum of the Philippines, Manila (Dr. Jesus Peralte).

The planks have the projecting lugs which are the hallmark of the construction technique. The lugs are 78cm apart, each about 30cm long, 16.5cm wide and raised 2.5cm above the plank, with four holes along their edges. To one of them a fragment of a transverse rib was tied down with rope made of palm fibre. Planks the whole length of the boat are carved out of a solid hardwood called doongon (1-teretiera litorales). The keel is simply a plank Scm thick, 45cm wide in the centre and tapering to a point at either end. The garboard strake (first plank) is 20cm wide and fi xed to the keel by hardwood dowels 19cm long, l.Scm thick spaced every 12cm. Round dowels of the other planks are 12cm long. The construction is therefore massive and strong.

Without doubt these archaeological finds establish that aspects of the. construction described by Fr. Alcina was used four hundred years previously. Important points still in doubt are whether such boats could be built with stone tools alone and whether these boats were the sea-going transports of a long-established culture trading with the mainland and within island Asia by Austronesian speaking peoples.

My own opinion is that the edge-pinning with long dowels is more conveniently done when a curved metal chisel (a spoon gouge) is available because the holes cannot be adjusted sideways a little with only a stone, shell or bamboo drill. The technical requirement for iron suggests that sewn boats predated dowelled boats and could explain why sewn boats persisted even in the Philippines. Presumably the people who spread across the Pacific did so before iron and the dowel technique had arrived. Dowel-built boats are known only where the metal spoon gouge was available, which was after about SOOBC in the Philippines (see note 42). In the Pacific to the east of the Philippines and Halmahera, many types of sewn boats built of planks were found at first 32 contact with Europeans, and their internal lugs, flexible ribs, and thwarts held together with lashings in tension, suggest that they were all originally related to the common lashed-lug tradition without the dowels. On that theory, those who migrated out of Island Southeast Asia before the arrival of iron did not take the dowel technology with them (see note 41).

33 I I

Figure 14

Excavation of a boat at Varrio Ambargo, Butuan, Philippines by City Engineer Procoso T. Gonzalez. The boat is about 2 metres deep in flood deposits. The edges of planks found here had round dowel holes about lcm in diameter, with transverse holes for the locking splinters through the dowels.

34 Boats in the Archipelago before 1500 AD

Accounts by early Western explorers give a picture that had presumably existed for centuries. They all stress that the way of life among the islands depended largely on boats. Travel, trade, migration, raiding and other social visiting were all done by sea and river. There is a sampling problem here, but even those driven inland by later arrivals usually have a maritime folklore.

Trade and fishing were the main supports of the boatbuilding yards. The trade among the islands included the collection of local products such as rice, cloth and sago, with valuables such as spices, sandal wood, ambergris, mother of pearl, and later dried shark fins and cave-swallows' nests for China. We know that they distributed rice, Chinese silks, trade ceramics, bronze gongs and swords. The archaeological remains of boats discussed in this paper could have been of similar local traders. There were also other larger trading boats, called jonques, which traded between , Malacca and China (Manguin 1979) and even visited Japan. Presumably the jonques also visited the main ports in the outer islands, such as T ernate, Banda, Larantuka in eastern Indonesia, and Cebu City in the Philippines. For speculations on the construction of the jonques see Horridge (1981: chapter 1).

Among the islands, heavy goods such as salt, rice, pots, iron, palm wine or gravestones were carried by boat, for most of the trading centres and boatbuilding regions did not grow their own food (as still is the case today). Fish provided a major source of protein. But the most conspicuous of the large boats were those adapted for carrying many men and used for raiding. Hunting parties went to other islands by boat, new lands were regularly colonized by boat; slaves were brought home at the paddles.

Chinese records quoted by Scott (1980) show that Visayan kora koras from the Philippines raided the Fukien coast with several hundred men, as witnessed by Governor Wang Ta-yu sometime between 1174 and 1189 AD. In the year 982 AD merchants from Mindoro were selling goods on the Canton coast, but the ships are not described. In the 13th and 14th centuries a port the Chinese called Ma-i on the southern coast of Mindoro was used by foreign merchant ships. From there cotton cloth was exported to mainland Malaya, as reported in the Tao I Chih Lueh of 1349 AD. In that period hundreds of thousands of pieces of Sung, Yuan and Ming porcelain were exported to the Philippine Islands, so much so that quantities of broken pieces are still being dug up. The Muslim ruler of a port, called Mao-li-wu in Chinese, visited China in 1405 AD and by the time the Spanish arrived the Muslim port city of Manila had boats trading as far afield as 35 Malacca and Timor. The Chinese said nothing about the construction but the known data summarized here leads to the conclusion that the lashed-lug boat was a standard design throughout that period. Nowadays we know of other styles that appear ancient, for example the of west Java and the lis-alia of Madura (Horridge, 1981) but we have negligible historical data on these before 1800 (see note 43).

Domestic trade among the islands was therefore well established by the twelfth century, and later the betrayal of peaceful agreements was often given as the major cause of war. From old accounts, we get an impression that the domestic economy of the maritime peoples of the Philippine Archipelago, as well as the people of Ternate, Banda, Ambon, Timor and , depended on continual trade in both luxury and staple products which were not produced locally or varied from time to time in supply. There was an extensive trade in the specialised products of particular crafts, such as iron, pottery, sarongs of silk or cotton, jewellery and dried foods. There were also slave markets and long-distance trade in luxury items. The large ports in the Philippines, stretching for miles on narrow strips along the coast, were often dependent on rice brought in by sea, and were local distribution centres for rice traded by sea. Scott (1980) gives many other examples of local trading of specialities among the Philippine Islands.

News also spread by boat. When Loaysa's ship the Santa Maria de la Victoria stopped in Liange Bay in 1527 AD, they were told that another ship had been wrecked to the south. It was the Santa Maria del Parra!, and one of the crew of the Parra!, captured by the natives, was soon told of the Victoria. Two other survivors of the Parral were taken to Maguindanao, where they heard from the Sultan of Brunei that the Portuguese Governor of Malacca was on the lookout for them. The Sultan provided transport for their journey. Magellan's men after his death found a man on the island of who knew enough to act as an interpreter, and near Mindanao they captured a Mora man who had been at Ternate with Francisco Serr'fo.

The kora kora was a ship of war for carrying men on raids for plunder or for slaves. Slavery in the eastern islands was quite different from that to which Spanish gentlemen, even members of the church, were accustomed in the 16th century. Members of one's own tribe in slavery for de bt were well treated, but the seagoing sultanates of the islands all captured and traded other slaves, mainly for their own labour and f ighting forces. Often they could win their freedom but some would be bought for human sacrifice. Raiding for slaves was an honourable y of making a living, and t he kora kora was needed for defe nce against raids as well as for forays. Coastal communities 36 would reckon their strength for attack or defence by the numbers of kora kora they had on immediate call. In case of sudden attack, the crew could pick up their boat by the outriggers and run down (or up!) the beach in a few moments. There were therefore two main activities based on boats, peaceful trade requiring cargo capacity, and raiding, for which the need was maximum manpower. Later, up to the late nineteenth century, there were many types of oared and paddled long narrow pirate boats, evolving by loss of the outriggers (Fig. 15; Horridge 1981, Figs. 3, 4).

Significance for the early European colonists

For generations, the boatbuilders, who formed closely knit communities, had supplied ships for local trade, overseas trade at least to Malacca, and raiding, before they were taken over by the new colonial empires. It must also be remembered that if it had not been for the local boatbuilding industry the new conquerors would never have been able to refit to get home with their spices and treasure nor would they have been able later to build up their colonial navies.

Concerning the use of local vessels by colonists, I cannot do better than quote Professor Scott's vivid writing: "The ship that Magellan's friend Francisco Serr?l'o ran aground in 1511 had been purchased in Banda, and when Governor Gonzalo Pereira sent some Portuguese officers to threaten Legazpi in Cebu in 1567, they made the trip from T ernate in two karakoas. So, too, when Spanish Governor Acuna attacked the Dutch in that island forty years later, they escaped with their families in four joangas, a sort of king-sized karakoa (see note 33). Legazpi himself used Filipino-built and Filipino­ manned vessels for exploring the , and sent Martin de Goiti to Luzon with fifteen of them in 1570. Smaller craft called biroko formed part of the Spanish fleet sent to Borneo in 1579, and more of them were stationed below Manila during a Japanese threat in 1592. Morga regularly used karakoas as dispatch boats and tenders, and by 1609 they were in such common use the Spanish King was moved to order some improvement in their design to protect their overworked indio crews from inclement weather. In the 17th century, whole fleets of them were being built to fight fire with fire as Mora contested Spanish control of the Visayas. All the major naval engagements joined during Sultan Kudarat's long lifetime were fought by opposing fleets of plank-built men-a-war with a thousand year pedigree. And the six boats Captain Francisco de Atienza carried up to Marawi in pieces in 1639 and reassembled on were genuine Philippine pre-fabs, too". (Scott, 1980).

37 Figure 15

An early 19th century pirate kora kora from North Borneo, showing the traditional features; two banks of oars, tripod mast, quarter rudders, square sail and upturned sharp prow and stern. On the right is a section of the same showing the lack of outriggers and the two levels for the men. (From a 19th century manuscript in the Museo Navale, Madrid, by Monle'on).

38 In his account of Francis Drake's circumnavigation, Thomas Cavendish mentioned a being built in 1587 for the Pacific crossing in a shipyard near the town of Arevalo in the central Philippines. The Governor of the Philippines, Gomes Per~s, went out with an expedition against Ternate in the Moluccas in 1593, taking with him 900 Spaniards and 200 ships filled with fighting men, including galleys (see below), galliots, , birreyes and others. In 1596 Esteban Rodriguez took another fleet against Mindanao, including, in addition to the above, many barangayes. The boatbuilding resources behind these operations can only be guessed at, particularly as the natives did most of the work. One of Fr. Alcina's duties was to supervise the shipyards where again the natives did the work. It is interesting to speculate how soon there was a transfer of technology, in particular to the use of stiff ribs fixed in by treenails, which allowed the hulls to be caulked. The use of the stern rudder and western oars worked on tholes, were already established in 1668 in the Spanish colony.

A note on galleys

A galley with oars, although prestigious because introduced by the colonists, and faster for the same number of men, was not so well adapted to the coral reefs, estuarine mud­ flats and mangrove fringes as the kora kora. The paddle is handier than the oar except in open water; the kora kora was designed to be picked up by the crew; the ram is not a useful weapon against a swarm of amphibious warriors on a flexible kora kora with outriggers. As a design for these islands, the western galley must have been a comparative failure but over a period of 200 years many were made, mostly diremes, for the Portuguese and Spanish, and some for the Indonesian Sultans (Nooteboom, 1951). A Sultan's galley called a buanga, seen by Pagee in 1767 is illustrated by Paris (1882-92, plate 331) with 100 men at oars facing back and 50 paddlers facing forwards. Perhaps buanga is the same word as joanga (see note 33).

Great care is necessary in interpreting names of ship types, especially galleys. Clearly many of the early mentions of galleys refer to the kora kora, which was the local boat for warfare over a huge area from Luzon to Timor. Those seen about 1598 by John Davis, an Englishman on the second Dutch Expedition, were kora kora. "He (Aladin, the sultan of Atjeh), hath very many Gallies, I thinke an hundred, some that will carry foure hundred men made like a Wherrie, very long and open, without Decke, Forecastle, Chase or any upper building. Their oares are like shovels of four foote long, which they use only with the hand, not resting them upon the Galley". (Markham, 1880).

39 At the same time there is evidence that when the Turks became influential in Atjeh (north ) after the fall of Malacca in 1511 AD, they showed the local carpenters how to build proper galleys with oars, and that these were numbered among the "galleys" which the Sultan of Atjeh took to Malacca in 1568. This is not surprising as the ships and trade of Atjeh were entirely oriented towards the trade with Malabar and even as far as the Red Sea (Nooteboom 1951).

Galleys with oars also were locally made further east, in imitation of those introduced by the Portuguese to Ternate and Timor. For example, a Makassar MS of about 1700 AD (British Library A. 12.360) describes how the ruler of Gowa (south Sulawesi) in 1626 had nine galleys fitted with a central rudder on the stem and also lateral rudders. They were long thin ships called gale, rowed with oars. Notwithstanding that Turkish galleys and Portuguese galleys were copied, the 8th century Borobodur carvings alone show that the kora kora was built for centuries long before the western medieval galley was ever seen in the East. At present there is some evidence that the Borobodur ships, however, and others of the first millennium AD, were influenced by other cultures with boats on the Indian Ocean (see note 43).

40 Two valuable models

In my searches in the museums of Europe I have found only two models which illustrate well the principles of boatbuilding described by Alcina. They both look like late nineteenth century models but come from regions that were isolated from western influences, and retain the essential elements of the old traditional design. These models are valuable and unique: both have come to light since I wrote Monograph 38 in this series.

The Berlin kora kora

The first is a kora kora without outriggers in the Museum fUr Volkerkunde, Berlin, No. iC 34355, from Sangir (Fig. 16). It is a long thin boat with six sets of ribs and lugs. The planks are held together edge to edge by dowels and the plank pattern (Fig. 22) is another variant of the now familiar form. A curious feature is that the ribs are set in place flat against the planks at the side of the rows of lugs, not over them, and there are lashings between the corresponding lugs on adjacent planks. It does not look as if the model has been repaired incorrectly long ago, but the above mentioned details of the lashings do not make good engineering sense because the load is not along the direction of the fibres. The thwarts are typical: two of them project through the sides as carrying handles. The line of the keel is gracefully curved and carried up at either end to rising stem and stern posts with supporting pieces at each side. Along the centre of the boat is a raised platform, the txrutlan in Alcina's account, typical of the kora kora everywhere (e.g. Freycinet, 1828, plate 37; Horridge, 1978, Fig. 10). We know from many accounts that the aristocracy sat here, usually with a gong orchestra, and here the fighting men were crowded on a raiding party. Besides the ribs, a detail that suggests a western influence is that the two masts are both single sticks supported in the western style, and one stands upon a rib. The original kora kora all had demountable tripod masts on pivots. Setting aside the question of the masts, we have here the most carefully built kora kora model still surviving; the originals themselves have now all disappeared.

The Breda prahu

Another revealing model, this time from Tanimbar, was acquired in 1962 by the Volkenkundig Museum Justinius van Nassau at Breda, Netherlands (No. 3600-4304). It was part of a collection of a Mr. Drabbe, and its age is not known. The boat is wide and rather flat with a hull similar to an old prahu of south Sulawesi (Herridge, 41 •

Figure 17

Prahu model from Tanimbar but constructed by a master boatbuilder of the tradition of south Sulawesi. This is model 3600-4304 of the Volkenkundig Museum, Breda, Netherlands.

44 I have previously illustrated other models in which the projecting lugs have been lost but numerous thwarts remain. One is Leiden model 37/582 which has a hull and plank pattern, without ribs but with numerous transverse thwarts (Herridge, 1978, Fig. 14). Another is Leiden model 254/28 which has numerous transverse thwarts and also fixed ribs (ibid, Fig. 16). These models illustrate the changes towards western hull designs at colonial centres in Indonesia.

Similar intermediate types survived in isolated places until recently, for example the Sumba prahu for conveying gravestones described by Nooteboom (Herridge, 1978, Fig. 15). The loss of the lugs, then of the thwarts certainly occurred in the 16th and 17th centuries in the Philippines at centres such as Manila, Iloilo and Cebu City when the Spanish taught the Visayan boatbuilders to fit stiff ribs permanently in place with treenails. In the period 1600-1660 AD the same change presumably happened further south under the strong influence of the Portuguese at Larantuka (East Flores), Timor and Makassar. For a picture of the times, see Boxer (1967). The more recent loss of the lugs in the Kai and Aru islands has been described by Herridge (1978).

The Makassarese jonks of 1660 AD sketched by Waldemar (1660) look as if they must have had stiff ribs. They were clearly modelled on the contemporary western ships, with a high stern (ambeng) built upon the original stern, and with a notch behind the bows. They were used for conveying rice, sandalwood and other heavy goods. These Makassarese jonks are the forbears of the of the Bugis in the nineteenth century; they are quite different in style and perhaps in hull structure from the jonques of Java (Manguin 1979, Herridge 1981, Chapter 1). In 1797 on the coast north of Makassar, Woodward (1805) mentions that the local prahus had ribs, but makes no further comment. We must await archaeological finds to follow the exact adoption of stiff ribs but it seems reasonable to accept that all ships of this region were built shell first with dowels in the plank edges.

45 r

The boat of Botel Tobago

To reach this island to the south-west of Taiwan one takes the bus to T'ai Tung then passenger ferry or aeroplane to Botel Tobago. The Austronesian speaking people there, relatively isolated over a long period but in contact with Luzon by boat, still make planked boats that are held together by internally projecting lugs that are lashed to the skeleton of the boat. As Hornell (1936b) said "there is no longer any possible doubt that the boat of Botel T onaga, the mon (of the ), and the Moluccan orembai all belong to the same class of naval design". The mon and orembai were long graceful canoes built of thin planks with projecting lugs inside. The mon was usually sewn, the orembai dowelled.

In the 1930's there were two types of boat on Botel Tobago; the smaller, represented by a specimen in the Hamburg Museum fUr Volkerkunde (No. A4579) was only about 3m long, with three strakes on each side of a narrow keel (Fig. !BA). The two ends are similar: the keel is spliced into upwardly curving narrow stem and stern pieces. On each side of these is an upwardly curving endpiece which is continuous with the wash­ strakes. The planks are fitted together edge-to-edge with dowels. All the pieces are cut from the solid wood, the various planks with their projecting lugs in place. These lugs are of only three types; the row across the centre of the boat are lashed to a U­ shaped frame (Fig. !Bb), others are lashed to a triangular bulkhead board near each end, and the third type holds the converging sides at each end of the boat (Fig. !Be). On the central row of lugs the rotan lashings pass through holes in the central rib. On the outside of the boat is a painted traditional pattern (which is seen in some Indonesian ilcat weaves) and there are multiple oculi at each end.

An example of this small type of boat, called the tatara (= boat) drawn for me in 1978 in Botel Tobago by Dr. C. Gut, is almost identical with Hornell's description. The plank pattern (Fig, !Bd) is similar to those found in traditional plank boats of the Kai Islands, the Moluccas and south Sulawesi, as described in other papers of this series. Probably the Visayans had similar plank patterns. As explained elsewhere the necessity of cutting the planks from the solid with a curvature in 2 planes, and lining up the internal lugs at the same time, together with respect for tradition ensured a constant plank pattern over long periods (Herridge, 1979, Fig. 37 and p. 49).

46 Figure 18 a) The tatara of Botel Tobago. b) The attachment of the central rib, with rotan binding passing through it from lugs which extend on both sides. c) The way of closing the ends, with a triangular bulwark lashed on one side to lugs, and by lashings which hold together the two sides to the thin stem piece. The triangular seat is not structural (a-c) from Hornell (1936). d) The pattern of planks and lugs; (d) and (e) were copied by Dr. C. Gut from a boat on the beach in 1978. e) The plank pattern of the end of the boat.

47 A second type of Botel Tobago planked boat, the chinedkulan described by Leach (1937) is similar in general plan, but larger than the tatara, with one more plank on each side and two ribs. The additional plank is attached only by dowels and lashings to lugs (which Leach calls comb cleats) at each end. In this boat the lugs are in a row along only one side of the ribs. There are five transverse thwarts as supports for seats. These boats are now rowed, and have wooden rowlocks.

According to Leach, similar boats were made by the Batans of the Northern Philippines and probably by the Ami people of Eastern Taiwan. In the 1930's there were traditions of journeys across the intervening 50 miles of ocean in the chinedkulan, but not in the tatara, which is too small.

48 The Whaling Pledangs of Lamalerap

Import records show that spermaceti and ambergris have been brought into China for at least two thousand years. Rumphius (1741; Chapter 35; p. 255) writing about 1690, reports plenty of whales around Ambon, but he cites only the neighbourhood of Timor as the source of ambergris and the sperm whale. Rumphius tells how the Timorese used it for sealing boats until the Bugis told them its value, whereupon they scraped it off again for sale. This sounds an unlikely story in view of the long history of trade in whale products.

Only in the early 19th century when the American explored that area, the route of the sperm whale through eastern Indonesia became known. Mussenbroek, governor of Ambon, gave an account in 1877. The sperm whales pass between the Indian and Pacific oceans through the straits at the eastern Lesser Sunda Islands, past the west end of and they can be found again at T ernate and off Morotai.

Two whaling villages still survive: they are Lamakera, at the east end of the island of Solar, and Lamalerap on the isolated south coast of Lomblen (Lembata). Their existence has long been recorded. The whaling boats of Lamalerap, called Pledang, were described by Weber (1902, 1923), who actually brought a model, since lost, to Holland. The way of life of the whaling village was described by Vatter (1932) and again by Barnes (an anthropologist) and others who visited Lamalerap since 1970 (see note 35).

0 0 Lamalerap at 8 34'5; 123 25'W, is reached from the airport at Ende, east Flores, by minibus to Maumere (12 hours), then to Larantuka (15 hours), followed by a 12 hour trip in a prahu or motor boat. The village is tucked into the exposed south coast of Lomblen where a short beach fills a break in the rocks. Along the top of the beach about 30 boathouses fill the whole length of the sand (Figure 19).

The village depends on fishing. All the houses have gardens but they obtain their staple foods, vegetables, fruits, maize, sweet potato and some hill rice by trading meat and fish with villages inland. They specialize in harpoon fishing but also troll for mackerel and tuna. About 1974 they obtained their first nets from an F AO fisheries project at Larantuka but I saw none in 1980. They hunt particularly for dolphins, porpoises and whales.

49 Figure 19

The beach and boathouses at Lamalerap, looking east at dawn along the ocean-facing south coast of Lomblen, Straits of Timor. Along the top of the beach is a line of boathouses (naje) each containing a whaling boat (pledang). Other smaller fishing boats stored between the pledangs, are dragged down to the sea every day at dawn.

50 The people of Lamalerap were partly immigrants. Their dialect is different from that of their neighbours. The ritual office of the "Lord of the Land" is in the hands of another village, on the hills above, which presumably had prior claim to the land. There is a local tradition that some came from Sulawesi 400 years ago, but their dialect is similar to that on the north coast of Lomblen, Solar and east Adonara. One clan supposedly came from Lamakera on Solar, another clan is called Ata Kai (men of the Kai Islands). In turn some men of the other whaling village, Lamakera, are said to originate from Ceram. We are not dealing with a long-isolated indigenous community, but with a mixed group with excellent trading traditions over a long period. It is significant that the surrounding coastal inhabitants of the eastern Lesser Sunda Islands built splendid outrigger canoes but not planked boats. Most of the ethnic groups of this area live inland and will purchase fish but are not otherwise concerned with the sea. The people of Lamalerap are now Catholics but retain animist traditions concerned with fishing, the sea and boats. Villages inland have lost almost all their rituals and become Catholic, but the traditional boats have not been a concern of the Church.

This is not the place for a description of this remarkable way of life, but I will give sufficient detail to show that the design of the boats is effective. The boats, called pledang or peledang are about 10m long, with 14 crewmen for a full complement. Each space in the boat, between one thwart and the next has a special name. Basically it is a pajala hull in shape (see Horridge 1979) with short outrigger booms and floats that are close to the sides of the boat. On the bows (Figure 20) is a bamboo platform with a structure similar but larger and stronger than the forward extension on the local dug­ out canoes. As shown in figure 22, the plank pattern of the hull and way of closing the bows are clearly related to the traditional designs elsewhere.

The amazing feature of these boats is that when you look inside you see the traditional lashed-lug design in one of its perfect forms (Figures 9, 21). The lugs on the planks are lashed to curved 'ribs' which are really tension bars, because each of their ends is drawn up by a twisted rotan lashing to the corresponding thwart that runs right across the sides of the boat (Figure 2d). Every one of the 30 or so pledangs is made in exactly the same way, although there are many boatbuilders in Lamalerap. There are 10 sets of thwarts, of which the third from each end are the stout ones that carry the outrigger floats. At sea the two sticks of the biped mast are lashed not very firmly to the ends of the fourth thwart. The small sail is the tilted rectangular rig at one of the limits of its distribution. The boat depends mainly on the oarsmen. The huge hawser of thicker than a man's arm, which attaches the boat to the harpooned whale, is kept in the 7th and 8th compartments and tied firmly to the thwarts next to the keel. 51 Figure 20

A pledang in its shelter.

52 The harpooner stands on the platform (hama). The crew pull the boat as close as they can to the whale. The harpooner launches himself with his harpoon using a weight to drive it into the whale. The harpoon is a large bamboo with an iron blade, forged in the village with a huge barb. The blade is fixed by its middle to the coir hawser and to the end of the bamboo so that it readily turns and gets crosswise inside the whale.

The whale dives and may submerge the boat completely. The hull is therefore dragged underwater with enormous force behind the whale. Usually several boats are attached to the whale before it is finally killed. Thinner harpoon lines are used for porpoise, manta rays and sharks. When a whale is caught the sharks that are attracted are also harpooned. They catch mainly small sperm whale (Physeter catodon) and some pilot whales (Globicephala macrorhyncha).

The rig is relatively weak and ineffective, partly because they rely on the oars when after a whale, but mainly because the mast and sail are expendible. If the whale drags the boat underwater the backstay breaks and the biped mas~ pivots loosely. The sail is too small for a boat of this size but that is all that is required. The tilted rectangular rig, normally called layar tanja in Indonesian, is peculiar to these boats and is called laya pledang. The fishing boats of Lamalerap and the surrounding islands have boomed triangular sails.

The pledangs show that the boats represented in the models, those described by Alcina and those dug up from beiow the flooded twelfth century midden, could have been effective, watertight and remarkably strong boats which men could confidently trust at sea. This ancient design has proved effective even when regularly dragged along as a sea­ anchor by wounded whales.

Reviewing the evidence, the pledang is not peculiar to Lamalerap and it cannot be explained as an importation from south Sulawesi a few centuries ago, as one local myth supposes. Boats of this common design, conforming to the engineering principles outlined, must have been widespread perhaps as far back as 500 BC. The Visayan boats described by Alcina illustrate a widespread design. Later in the 18th and 19th centuries similar boats were made for sale and distributed by Kai Islanders (Wallace 1869), carried by migrants or refugees from other islands and spread by Sulu raiders on the look out for slaves. Many survived into this century in isolated islands (Hornell 1920) and we are fortunate that a few examples survive intact as the pledang of Lomblen and the prahu belang of the Aru Islands (Horridge, 1978, Fig. 19) to corroborate Alcina's accounts today. As yet there is no example of such a boat in a museum anywhere in the world. 53 The Lashed-Lug Construction Technique

The Contribution of James Hornell

''In building, the planks are first secured together with vertical pegs; afterwards when the hull is complete, frames or ribs are fitted over the vertical rows of projecting cleats and tied thereto by cords made by black palm fibre. Thus the ribs do not lie against the inner surface of the planking, but are separated therefrom by the thickness of the cleats to which they are tied. By this device no metal fastenings are required; such a hull possessed great elasticity and stands bumping in the surf in a way that no metal-fastened boat would long survive". (Hornell 1920; 59).

Hornell so described the lugs and lashings, and he was the best of many who remarked on this method of construction. He was describing plank-built boats with outriggers, which look like cargo-carrying descendants of the kora kora, from Halmahera and Batjan (Bacan) Islands (Fig. 10c). Subsequently Hornell (1935, 1946) described the comparable details on the Moluccan orembai, which was a long graceful passenger boat with upturned ends. My pwn conclusions, though based on more diverse evidence which was not available to Hornell, differ little from his, but go further in establishing that the lashed-lug boat was a standard design with a wide distribution in the east, going back for at least the past thousand years; and very serviceable when well made.

For years Hornell (1935, 1946) stressed the remarkable similarity of the Moluccan boats to iron-age boats bearing projecting perforated lugs that have been dug from the bogs of Northern Europe, for example the Nydam oak boat and the Hj6rtsprung boat, illustrated among others by Greenhill (1976). Hornell also remarked on the similarity of bifid stems, winged stems and bailing dippers (Fig. 21). Two points can be added to his speculations. First, although little can be said without more evidence, the growth of maritime archaeology may well reveal lashed lug boats from ancient ports in between the two far-flung extremes where they are known so far. Secondly, the design and effectiveness of the surviving examples show one possible way how the lugs on the European boats may have been used, and give a good deal of insight into one ancient subtle solution to the problem of building a boat.

54 (b)

(c)

Figure 21

Details mentioned in the comparative study of early boats. a) The essential lashed-lug structure. There are ten of these ribs in a Lamalerap pledang. Compare figures Zc, d and 17. b) The design of bailer noted by Hornell (1935), to have a peculiar distribution. c) A winged stem (see also Horridge 1981 fig. 25).

55 The broad perspective in Island Southeast Asia

I have referred briefly to the wider implications of any study of ancient boatbuilding techniques in Island Southeast Asia, and should perhaps now spell out more definitely my own theory of the relations between the dug-out canoe and the two main t ypes of built-up canoe, i.e. sewn and edge-dowelled.

In the following I have relied on the broad archaeological perspective summarized by Bellwood (1978), in particular his dates for the metal age in southeast Asia (1000-5000 BC) and for the earlier spread of Lapita pottery by speakers of proto-Austronesian who passed east through about 1500 BC, to become .

First, I believe that the speakers of Prato-Austronesian owed their success to their boat technology. I assume that their pre-ceramic cultures had rafts and dug-out canoes and migrated in short stages. Over a long period before the metal age in , i.e. before about 1500-1000 BC there developed a more sophisticated maritime culture that made possible voyages in open water, out of sight of land, through the night, carrying fire, food, family, chickens, pigs, tubers and shoots. Voyages of were rapidly backed up by trade routes, as shown by the extensive and rapid movement of Lapita pottery in some regions, beads and obsidian cutting tools in others (Bellwood 1978, Chapters 7 and 8).

Taking into account all available evidence, which is mainly my reading of the comparative anatomy and distribution of canoe types of the Pacific and Indian Oceans (Haddon 1936, Hornell 1936, 1946) together with some knowledge of canoes and boats that are effective in warm seas. I suggest that this first wave o~ real voyages depended on a group of techniques (a) sewing additional planks on the dugout hull and being able to make sewn boats when large trees were not available, (b) understanding the proportions of a successful outrigger canoe and the properties of the materials used in construction, (c) strengthening the sewn planked hull with thwarts and ribs lashed to lugs and other internal fittings, of which there has been a great variety in the Pacific, (d) techniques for carrying the essentials of life at sea for some weeks, (e) navigational techniques and finally (f) perfecting a sea-going rig which was probably a triangular sail sewn to two booms, supported by a movable pole with a rope to the windward outrigger. The union of the two booms was attached to the front of the hull. These people had no pulley block, a point which may have limited the improvement of their rig designs. This rig was without a fixed mast, with a boomed triangular sail pivoted where the two booms meet, I have called the proto- rig, and it still persists in Madura (Horridge 56 1981) see note 44. Pacific boats subsequently evolved into a variety of sewn, dug-out, built-up or lashed but never dowelled canoes in the Pacific, including the mon of Melanesia (Hornell 1935, Plate 6) and the double canoes of . These people had no metal. They also spread to and . There is little evidence of skin boats.

After those voyagers had departed in the period 2000 to 1000 BC, the mainland cultures further developed with wet rice, metal and the buffalo. These mainland cultures also spread by similar built-up canoes and planked keels over Indonesia and the Philippines, trading with and driving out people already there. One of the many things they brought from mainland Asia was the technique of fixing the planks together by dowels, for which I suggest they needed a metal tool. This technique enabled fine planked boats to be built, but the dowel technique spread about as far as the knowledge of metal working and did not entirely replace sewing in backward places. These people used a different rig, with a tripod mast and tilted rectangular sail, of which the oldest record is at Borobodur (8th century AD), (see note 43).

Thirdly, a different boat technology, based on the bulkhead instead of the rib and thwart spread into Java from south China or Cambodia, probably in the period 300 AD to 800 AD, corresponding to the opening up of regular trade to China and the flourish of Hindu culture in Java. Possibly there were many types of mainland boats in this period.

In the fourth stage, at the arrival of the Portuguese and the Spanish, the boatyards in the main cen.tres were taken over and re-educated. The method of building a shell fixed with dowels persisted everywhere but the lashed-lug technology survived only in out-of­ the-way places. Stiff ribs fixed with treenails replaced the lugs, and the boats could now be caulked. The progressive introduction of further western features such as the single mast, stern rudder and deck has been catalogued elsewhere (Herridge 1979, 1981). These changes took place more rapidly and completely in the Philippines than elsewhere. The old styles persisted longer in Indonesia where the Dutch had little influence on boatyards in the outlying islands.

Relations to the West

We can now consider more clearly the resemblance to early boats of northern Europe.

The lugs, the bailer, t~e winged stem (Figure 21) and the bifid stems noted by Hornell (1935) could also have been known in the Indo-Aryan cultures of the Indus and Mesopotamia valleys sometime between 4000 and 2000 BC. These ideas could have 57 arrived in small boat form with speakers of the Sanskrit family of languages over long distances via the Danube and Volga trade routes from the Caspian and Black Seas into northern Europe about 3000-2000 BC. At that time so far as we know, the plank boats of that region were sewn. This admittedly highly speculative theory postulates the spread of an early Asian boat design that later developed along separate paths at opposite sides of Asia. For further details of lugs on Bronze Age boats in England, and a discussion from an entirely different position, see Wright (1976).

The stages after the sewn boat developed in different directions in regions around Asia. In northern Europe overlapping strakes came into use; in Egypt and spreading into the Mediterranean, the planks were fixed edge to edge with mortices and tenons: in China broad planks were nailed to bulkheads with iron; in Indonesia the traditional and only known method was to fix the planks edge to edge with internal dowels, but data is lacking before the 12th century.

The Indonesian round dowel technique was either a separate invention of the Austronesian-speaking seafarers, or it came from India: certainly it contrasts with the Chinese methods. Boats on the Indus in 2500 to 1500 BC could have had dowels between their plank edges, because the technique was known, as shown by their construction of wheels (Johnstone, 1980, pp. 179-180). The round dowel technique is an adaptation to tropical marine conditions, and it never spread to the West although many cultures including Persians, Phoenicians, Egyptians, Greeks, Romans and later the Turks, all built ships on eastern seas as well as on the Mediterranean. Perhaps the technique was not carried westwards because the eastern boatbuilding craftsmen did not cross to the Mediterranean coasts: rather, movement was the other way. It seems a reasonable conclusion that the dowel technique spread eastwards from India with the use of metal. This brings us back to the beginning of the metal age in southeast Asia, which moves out to the islands about 500 BC (Bellwood 1978, Chapter 8).

The conBe'P'flC88 of the lashed-lug technique

Building the hull of a planked boat was one of the most complicated of the craft skills, and boatbuilders were an important part of society. They had no plans or written instructions. For generation after generation they turned out hulls with perfectly fitting curved planks carved from solid hardwood. The evidence suggests that the basic construction principle, with tension on the lugs making the hull into an inverted compressed arch, never changed for more than a 1000 years.

58 The design has a combination of features that fit perfectly the materials available and the magnitude of the stresses encountered in a boat of this size. It illustrates the principle that the loads must be distributed to avoid stress concentrations and also the principle that a break at any point should allow the forces causing the break to be immediately redistributed among many other load-bearing structures, and be dissipated rather than encourage the fracture to spread. Goods of leather, basketwork or cloth also have this property. As explained in Figure 2, the lashings on the lugs turn the hull into an arch in compression, and these compression forces at the plank edges must be overcome before the planks begin to separate. The first signs of leaks can be corrected by tightening up the lashings, or adding additional twisted lashings. The dowels within the edges of the planks prevent them sliding relative to each other as the boat twists and bends between the waves. The shear forces between the planks are thereby not transmitted to the ends where they would burst the joint with the stem and sternpost. Compressing the edges of the planks against each other increases the friction. The whole hull becomes one rigid prestressed structure, like the moulded shell of a modern reinforced plastic hull.

In Indonesi~, until recent influence from Portuguese or Dutch colonists, planks were never bent by soaking, heating or steaming them, for such planks are no longer stable. The construction, all in grown and carved hardwood, gives the boat a longer life. In my experience, such boats built in Indonesia, show signs of opening seams and rotting planks only after about 10-15 years. They are then easily repaired. The locking pegs, if present, can be knocked out of the dowels and the whole boat can be dismantled for refitting the plank edges. A feature of the lashed-lug design is that it can be re­ assembled perfectly when the planks are slightly narrowed, whereas with fixed stiff ribs the holes for the treenails then come in the wrong places. Dowelled boats made entirely of hardwood last 30-40 years in the tropics and can be repaired indefinitely. Contrast this with the fate of Magellan's second ship, the Trinidad, which was so rotten after eight months in the Archipelago that her seams opened when she was loaded with spices at Tidore. In the tropics all the early western ships suffered the same fate when two or three years old because they were held together by iron fastenings and nails. The iron rots the wood in its vicinity, besides crumbling away itself. The Portuguese in India and the Spaniards in the Philippines soon after their arrival had ships built locally for service in the tropics. For the same reason the trade between Malacca, Java and south China had for centuries been conducted in foreign vessels and not by C hinese junks, which we re basically river boats fixed together by iron staples and rivets. Later, the European lndiamen we re better caulked and had no exposed iron.

59

A consequence of the way of building with the lugs in situ was that in a particular culture every plank in the boat had a constant shape and was known by name. This leaves plenty of scope for local variants, especially at the bows and stern (Figure 22). The whole boat is a structural unit and once the design and plank shapes are established it is very difficult to introduce a change in the ratios of lengths without destroying the whole pattern. A structure, once learnt, was stabilized in the culture as part of a strong oral tradition. In comparable boatyards of the present day in south Sulawesi, every detail of the traditions are strictly preserved by long apprenticeship and rituals that are based on a respect for ancestral methods. Even so, new ideas are incorporated into the tradition, usually by copying western styles.

Internal lugs are only one way of attaching planks to ribs in sewn boats, but the technique can be found far out into the Pacific. An example drawn by F.E. Paris (who sailed with d'Urville on the Astrolabe in 1826-29) is reproduced by Hornell (1936, Fig. 188b) with a Tongan vaka, and by Dodd (1972, page 146) with a Tongan tongiaki. The structure was so arranged that the plank edges could be compressed together. In another method, the stitches of the sewing were put through projecting ridges carved on the edges of the planks. There is no sharp dividing line between dug-out canoes built up with planks and boats built on a keel, but sewn boats usually have a broad flat keel.

Large sewn ships without internal lugs are also known, the earliest being the boat of Cheops (3000 BC) and the most spectacular the , with examples surviving to the present day. Some dhows also had internal dowels within the plank edges. Many examples of small sewn hulls without lugs or ribs can be cited from south India and from the Pacific (surveyed by Hornell 1946). There is an excellent nineteenth century model of a sewn Tagalog prahu from Luzon in the Museum of Ethnology, Rotterdam (Cat. No. 4950). It has an overhanging walkway of bamboo all round for numerous paddlers and was probably called a . Elsewhere in the Philippines, sewn boats were in use on the Cagayan River in Aparri (Northern Luzon) and among some turtle-eating minority groups off the southeast coat of Luzon in the 1920s. We do not know if they were an ancient design or secondary.

61 Conclusion

The history and spread of the Austronesian-speaking peoples is a story of many water­ borne cultures. We do not know how the earliest pre-ceramic cultures spread across the archipelagoes to Australia, New Guinea and the Pacific, but by 2500 BC early Austranesian speakers made long journeys by sea through Melanesia, as shown by the spread of their (Lapita) ceramics. Their boats were presumably double canoes with freeboard raised by sewn planks with internal lugs and thwarts: some could have been sewn planked flat-keeled boats. With an impressive array of techniques and traditions for migration over water, these people later became the Polynesians, mixing with others along the way.

About 500 BC, the use of iron and bronze spread from mainland Asia, together with the curved chisel that enabled planks to be fitted edge to edge with dowels. Progressively larger and more perfect boats were developed in the Philippines and Moluccas, using the lashed-lug and dowel method of construction. Together with planked ships from the mainland and from Java, these boats were used in the islands by many maritime kingdoms that depended on raiding and trading by sea. Extensive and persistent boatbuilding traditions grew up. In the Western literature these techniques of boatbuilding have not been adequately appreciated; their significance for the continual movement of peoples among Island Southeast Asia has been ignored; the stress patterns and principles of their design have not been elucidated; standard accounts omit the lashed-lug technique; the importance of the traditional local boatbuilding skills for the maintenance of the early western trading colonists has not been realized: previous accounts of primitive boats for the last two centuries have hardly mentioned the kora kora or karakoa, which was a highly efficient plank-built boat, able in its larger versions to carry hundreds of people in warm waters. Furthermore, no full size examples of these boats are anywhere on exhibit or in collections, and it is an urgent matter to preserve the very few types that are now almost extinct.

62 NOTES

1. Lugs as described in this paper as fi xi ng attachments in canoes or on planks are illustrated over a wide area by a variety of authors; for in Fig. 5 in Hornell, 1936, Vol. 1 (Polynesia, and ): for Solomon Islands in Fig. 62 in Haddon, 1937, Vol. 2 (Melanesia, Queensland and New Guinea). In his extensive work on the names of canoe parts Friederici (1912) comments as follows: FUr diese kleinen durchlocherten Osenleisten habe ich in den Molukken und auch sonst verschiedentlich keinen Namen finden kt5nnen". So he used a

widespread Melanesian n~me, patnati. At Bacan, Friederici found tambuku; at Galala, Ake Selake and Tobelo (Halmahera) maru maru; at Weeda Bay .( Halmahera) popula. At Lamalerap they are kelik. In a variety of dialects, notably Ambon, Banda, Butung, Baju at Ternate and Mandar - I have been given the word buku which is a general word for knot, knuckle or lump, as was presumably also the Visayan word tamboko recorded by Alcina. Sometimes the lugs survive although no longer used. Very significantly the south Sulawesi method of fixing the positions of the ribs in a pajala hull is to divide the length of the keel into about 12 tambuku separated by 13 ruas or ruang (Horridge, 1979). Tambuku and ruas are the nodes and internodes in a bamboo stem. They divided a length presumably by choosing a bamboo stem with nodes the right distance apart, or at least thought of the operation in that way.

The word for the flexible rib, not given by Alcina, is less difficult to trace to its root. At Lamalerap the rib is lulu; at Ternate Friederici (1912) found gilu; in Mandar and occasionally elsewhere I have been given kelu or kilu (= curved in Malay). In Madura one finds keting or kerting. In modern dialects the projecting stiff ribs in prahus are taju, meaning something sticking up. Besides these, two derived words, tulang (= bone) and gading (= elephant's tusk) are used, but understandably there is a dearth of dialect words for the large stiff ribs and floors of prahus which are inferred to be taken over from western designs.

2. Confusion is easily reached in discussing words for traditional drills. In Madura the tool held against the chest while a threaded block is drawn up and down a spiral shaft, is coarsely called locor (with Indonesian c as ch in church). A similar word in Malay, locok, meaning the same oscillatory motion, will often be interpreted as 'masturbate'. Nowadays most people only know the Dutch word bor or the Tamil gurgi for a modern twist drill or auger, and they have never heard of cutting round holes with a chisel. 63 From Alcina's description it is clear that the lokob was a different root, meaning a spoon drill or gouge of iron, perhaps similar to those used in medieval Europe for the construction of houses and ships. In Sulawesi, a prahu before it is launched must have a drain hole cut diagonally in its keel. The drainhole is the navel (pU88i) of the boat with great symbolic significance in its birth process. The puBSi should be cut with a special curved chisel (pa'otere in south Sulawesi) which is clearly the same tool as the Visayan lokob. In a few places an ancient curved chisel is still the prized possession of a master boatbuilder.

3. Binkung is the word now used by Buginese farmers for a back hoe and by boatbuilders of south Sulawesi for their standard square adze which is used for all purposes. In , a blikung or bli'ung is the curved adze for making a dug-out canoe. Many of the boatbuilders use other regional words (Herridge, 1979, p. 44).

4. Taking two planks only from a trunk was the standard method, as described for the Kai Islands by Wallace (1869) and for south Sulawesi by Woodward (1805). Possibly a slightly bent trunk yielded two curved planks that could be used on opposite sides of the boat.

5. In all of the boatbuilding cultures of offshore south-east Asia the individual planks had traditional names which survived into this century. The long planks could all be prefabricated by memory .to a rough shape in the forest from suitable trees. Every boat was built to the same pattern. Some plank names are listed by Friederici (1912) and by Herridge (1979). Most plank names have now been forgotten because nowadays planks can be shaped as the boat is built in the boatyard from a pile of timber.

6. It is impossible to cut planks from close-grained hardwood so that they curve in

at both ends to meet the stem and stern posts. To avoid this diffic~lty the traditional plank pattern in Indonesia, for as long as can be traced back in old models, has had short pieces where the ends of the plank turn sharply in towards the stem and stern (Herridge, 1978, 1979). The exact pattern is different for each boatyard, and therefore can be a useful feature for identifying origins, see Figure 22. In a few boatyards the Dutch introduced the bending of planks in the 19th century.

64 7. The practice of building a planked boat on the base of a dugout canoe was widespread among all the from the Indian Ocean to the Pacific. In mainland the canoe base was opened out by force and heat (Warington-Smyth, 1902). In south Sulawesi the word for keel, kalebiseang means "the original boat". In Madurese the word for keel, dasar, is also used for the hull of a canoe.

8. The use of the mortice and tenon can be traced among ancient highland societies

with fine traditions of building houses and granaries, e.g. the Toraja o~ central Sulawesi. In the non-hispanized society of Ifugao in northern Luzon, fine mortice and tenon work is still done with nothing more than a chisel-shaped adze. Polished stone (nephrite) adzes which could have been used for the same sort of work have been recovered from archaeological sites. The mortice and tenon joint for the extension of the keel is the focal point of the ceremony when the boat is 'conceived'. Symbolic objects are placed in the mortice, which represents the vagina. When the tenon, representing the penis, is fitted into place, the spirit of the new boat is considered to have been formed. For further details of the ceremonies and beliefs, see Herridge (1979).

9. All known traditional planked boats of the Philippines, Sulu, Sulawesi and the Moluccas had the keel extensions and also the stem and stemposts in a continuous smooth curve with the keel, forming a moon-shaped keel (lunas). Another type of traditional boat, represented by the Lis-alis and the Janggolan of Madura (E. Java) has two stems morticed in near the end of the keel, forming a flat-front to the bows and a fork with the keel (Herridge 1981). A single stem morticed at an angle into the end of the keel was probably an early western introduction in all of this area, and such hulls are now everywhere distinguished by the name sekoci (= , from the Dutch).

10. Alcina imputes all the strength to the transverse thwarts and does not describe the significance of the flexible ribs. He sees the planks as held together by the dowels and does not emphasise that the boat is held together by the tension in the rotan lashings (Fig. 2).

11. According to W.H. Scott, bahe is the wood of the anahawis palm. The use of certain tree species, Caesalpinia sappan and Mimusops elengi, for making dowels for the edges of boat planks was well established when the first western descriptions of these trees was made by Rumphius in Ambon about 1700 AD. 65 These timbers and mangrove wood are still in use today (Horridge, 1978, 1979). Brazil wood (C. sappan) came into medieval Europe as a source of red dye. The best sources are Luzon and Sumbawa according to Raffles (1819, 227). The Malay name is sepang from Sanskrit plttanga; sometimes it is called soga jawa (Sundanese). Exactly why certain timbers are best for dowels (hidden in the plank edges), or for treenails (which are exposed at their ends), has never been investigated.

12. The two sets of corresponding holes on the adjacent planks are made coincident either by first marking a line on both planks together, then boring them, or alternatively by boring one side first and marking the places upon the other plank with white chalk dust which is placed on a wad in the hole. A perfect match is rarely obtained, so the spoon gouge is used to widen one side of a hole. Usually the dowels tighten up as the planks are driven together. In Malaysia the dowels are thinner and closer together. A modern degenerate technique, spreading from into Indonesia, is to knock in nails of a non-rusting alloy with their heads clipped off. The dowels counteract sheer forces: they are insufficient to hold the boat together.

13. Such tongues at stem and stern posts are not a feature of any traditional design that I have seen, and it is possible that Alcina was referring to the simple re bate in the stem, called in Sulawesi boatyards tungkulu.

14. In most boatyards, I have not seen dowels cross-locked as Alcina describes, but I did not look out for that detail until awakened to the point in 1978. Bugis and Makassarese boats I have examined do not have them, but the locking pins are found in Madura (where they are called ba'tun) and in Bali where I was given a general Indonesian term, paku pengancing, meaning fixing pegs.

15. The fact that these flexible branches have a small range of diameters for a wide range of boat sizes can perhaps be taken to indicate that the branch must be thin

enough ~o bend but sufficiently thick to carry all the tension that can be exerted by the rotan bindings to the lugs. The design in figures 2(b) and 8 requires a more flexible branch than that in figures 9 and 17. In the Hjortspring boat, Scandinavia 300 BC, alder branches were used, and alder is flexible when green. The significance of this detail has been missed in the past.

66 16. Holes in the flexible ribs are not mentioned in other accounts, and in my observations of models and real boats, holes occur only at the ends of ribs (Fig. 9).

17. The word batangan for the outrigger boom, together with some other words, links several of the boatbuilding peoples of Indonesia. Outrigger boom is baratang in Mandar Macassarese and Bugis, baratahana in Bahasa Baja of north Sulawesi, baratano in Butungese and barajlnJBn along the north coast of Madura. In the Aru Islands lalangan means a thwart inside a boat.

18. The tadi is the flexible connector piece between outrigger boom and outrigger float in the Mandar language of west Sulawesi. The word is tadi in Mandar, cadi in the Malacca Straights and cedek in Bali.

19. Kate in Visayan is undoubtedly the same as kater in Madurese and katir in Balinese for outrigger float. In Java, perhaps by an independent route, the float is kicak. Through much of Polynesia the word is kiato or giato but means outrigger boom there (Friederici, 1912).

20. The terms attributed to Galv~o in this particular quotation are typically Moluccan and are related to words that are widespread today in the Moluccas in contrast to words in the Philippines and Madura. Outrigger boom is today ngaju at T ernate, naju at Tidore; the outrigger float is sema1 at T ernate, sema sema at Tidore; and the flexible connector piece between these two is the pagu at both places.

From broad study of the names of the principal parts of outrigger canoes, Friederici (1912) concluded that the technical ability to build outrigger canoes was carried outwards over the Pacific from Indonesia with the original migrations. He traced one route from the Philippines southwards then eastwards through the Solomon Islands and another route directly eastwards from the Moluccas. The word saman for float is characteristic of the latter route, and for all groups of people who are distinctly Malayan: on the other hand, the word kati, rakit, kiato, giato, iato is from the Philippine origin and is characteristic of all distinctly Polynesian groups, with a curious offshoot in Madura and Bali. The design of canoe bows shaped like a fish jaws, as in the Mora and Sulu (Hornell 1920), the Madurese and Balinese jukoog and the sekong of east Java (Nooteboom, 1932) is linked with the katir group of names.

67 21. Many western historians have considered these paddled "galleys" to the east of Malacca to be copies of Mediterranean galleys of the time introduced by the early colonists (Nooteboom, 1951). In fact Turkish rowed galleys were masters of the north-western Indian Ocean when the Portuguese arrived, and may have ventured east of Malacca. Knowledge of rowed galleys would have been commonplace in the Arab world of 1500 AD, and a little before that could have been carried to the Philippines along with Islam. But, considering the construction technique, and the adaptation to local conditions, the only reasonable conclusion is that the kora kora was an indigenous design that had been used for centuries before rowed galleys were ever heard of. The Portuguese introduced their own galleys after 1500 AD with a ram at the bows and men at oars in the Venetian style, and they used numbers of them in their colonial wars.

22. Seatia = pinque; a Mediterranean vessel with two masts, two large loose-footed lateen sails and a very small mizzen. Alcina is saying here that the relation between the mast, the halyard and the sail on their tiled rectangular sails is

~imilar to that on the western lateen rig. The sails themselves are quite different. The similarity is that in both types the sail pivots and swings in any direction at the top of the mast, so that it can be adjusted to go into the wind with the sail along the boat behind the mast, or downwind with the sail in front of the mast across the boat. To tack, both types had to wear ship and bring the sail and sheet around the front of the mast. For distribution and detail see Horridge (1981); for performance data of this sail type on a Sulu vinta, see Doran (1972).

23. The Lutaso, or Lutaya, were Moros, probably Samal-speaking, whom the Spaniards in the 17th century considered as their worst enemies: they were at least in part vassals of Maguindanao (what is now called Cotabato on Mindanao) and sometimes allied also with Jolo and . Their boats were from 12 to 22 metres long, carrying crews of 60 and upwards. The Spaniards thought, rightly or wrongly, that the Lutaya were newcomers, but the term has long since disappeared. Early accounts in Blair and Robertson (1903-9) make them out to be boat-dwelling fishermen, and all references make them out to be seafarers, possibly sea-gypsies (Orang laut or Baju).

24. This rig is the layar tanja of Indonesia (Horridge 1981; Fig. 1 and Chapter 3). In Indonesia, beautifully soft sails were woven from fibres of the Arenga palm. In 68 fact sailmaking was influenced by the distribution of the Arenga (sugar palm) which grows in dry areas and was transplanted to places like Penang for the benefit of the colonists, even into the eighteenth century.

25. Some beliefs and ceremonies connected with boatbuilding in south Sulawesi are described by Herridge (1979). The sacrifice still persists but is nowadays a goat.

26. Increasing the freeboard with plaited mats is an ancient and still common practice in South East and East Asia, for example in Solar and Lomblen near the Straits of Timor. The patterns of plaited mats have even become stylized as carving on the gunwales.

27. There is doubt whether any of them rowed in the European style before the westerners arrived, but long oars were used in Turkish galleys in the Indian Ocean before 1500 AD. They are illustrated for example on the Portuguese chart of 1519 by Homen-Reineis in the collection of Armando Cortesao, published in 1960.

28. Paddles of this type were also used on large boats in Indonesia up to the nineteenth century.

29. For much of the nineteenth century, piracy was the main activity of the last independent sultanate, in Sulu. Their kora kora without outriggers easily outmanoeuvred a in light winds. The pillage by karakao was not suppressed until after the advent of the steamship.

30. In the Moluccas (Ternate) the bangko was a plank-built boat with paddles and sail for passengers, with heavy outriggers of four crossbeams. In remote islands around Sulawesi such boats still exist for transport of cattle and heavy goods. In the Philippines today, a barato is a small fishing boat, usually an outrigger canoe and a banka is a larger boat that may have outriggers (Spoehr 1980). Bangka or wangka is the Austronesian root wood for boat.

31. A league was probably about 4 km.

32. Alcina's statement confirms that the central stern rudder was copied from the colonists. We find pottery revealing continuous contact with China for centuries, but as in Indonesia, little sign of Chinese influence on boat technology, 69 possibly because the Chinese used foreign ships for their trade. Anyway, the Chinese deep stern rudder was unsuitable for shallow seas with reefs and currents: also it was not appropriate for the smoothly curving stern of local boats.

33. During the first half of the 17th century the Spaniards seem to regard a joanga as a large karakoa, even distinguishing by the number of oars, e.g. 50-100 for the latter, 120-130 for the former. The Spaniards had both kinds built to fight at sea. In the MS, the caption on the drawing (Fig. 13) says "Las caracoas son como las joangas exceptas los rodas de papa i ", and it's obvious that the embellishments are strictly Spanish. The word duanga is used for "boat" by the Toraja of Central Celebes.

34. "The Journal or Daily Register, containing a true manifestation and historical declaration of the (first) voyage accomplished by eight ships of Amsterdam who sailed the first day of March 1598". Published, London 1601. A facsimile of the copy in the Prins Hendrik Maritime Museum was published in 1954 by Walter J. Johnson, Norwood, New Jersey and Theatrum Orbis T errarum, Amsterdam.

This volume is not to be confused with the later one by Jacob van Neck, "Journal van de tweede reys naar Oost-Indien", 1600; publ. in Amsterdam, which relates the second expedition of Jacob Cornelius Neck. The standard accounts of the Dutch voyages, in Dutch, are to be found in the long series of volumes published by the Linschoten Vereniging, see note 37.

35. Three reports on World Wildlife Project 1428 "Cetecean fishery at Lamalera, Lembata" have been published. R.H. Barnes, (January 31, 1980) Cetaceans and cetacean hunting - Lamalera, Indonesia: 82 pages. E.D. Hembree, (August 1, 1980) Biological aspects of the cetacean fishery at Lamalera, Lembata: 55 pages text, 30 pages tables. S. Silalahi, (October 30, 1979) Observation and research on the cetaceans fishery at Lembata, Indonesia: 5 pages. These reports can be obtained from the IUCN/WWF Project Officer for Indonesia at IUCN, Avenue du Mont Blanc, CH 1196 Gland, Switzerland. The great significance of the boat structure has been entirely missed by the anthropologists and ecologists.

70 36. Useful early Spanish accounts are those of (a) Francisco Combes, Historia de Mindanao y Jolo, Madrid 1667, republished Oablo Pastells & W.E. Retana, Madrid 1897. (b) A~tonio Galvao, Tratado dos descobrimentoa antigos e modemos (1563) republished Oporto 1731. (c) Translations of numerous early papers in Blair, E.H. & Robertson, J.A. (1903-9).

37. Useful Dutch accounts are in the long series of volumes published by the Linschoten Vereniging published Nyhoff, s'Gravenhage, Netherlands. They are similar to the English Hakluyt series.

38. I hear from Janet Wisseman that Javanese inscriptions reveal extensive maritime trading from the north coast of Java back to the lOth century, with many ship types including , kora kora, , mayang, prahu pukat (pukat = various kinds of nets to put around a fish shoal). We can only suppose that the names had the same meanings then as after 1500 AD.

39. On some prahus today the only pulley is at the masthead, and some have only a hole or a bar for the main halyard. The other early rig of this region, the protolateen rig with the boomed triangular sail, requires no main halyard because the upper boom is pushed up by a loose pole (Horridge 1981, Chapter 7). The limit of the spread of the pulley block from mainland Asia in prehistoric times is not known, and I have seen no discussion of indigenous Pacific rigs as being limited by a lack of the pulley, but it appears from illustrations that indeed they were so limited.

40. Doran (1972) made sailing tests on a wa from Puluwat in the Carolines with single outrigger and boomed triangular sail (protolateen or oceanic lateen rig), a Sula vinta (with double outrigger and tilted rectangular sail) of about the same size and sail area, and a modern Bermuda rigged . He found that the 0 trimaran was superior up to 68 to the true wind, the triangular sail was superior 0 0 0 0 from 70 to 96 , and the rectangular sail from 96 to 180 • His measure of performance was (boat velocity)/(true wind velocity). This ratio reached 0.65 in 0 0 the wa around 85 and in the vinta around 110 (see Doran 1981, and note 44).

41. There is considerable doubt about the exact distribution of the dowel technique, not only for past centuries, but even for existing canoes in museums and remote places, because the dowels are invisible. One particularly interesting example 0 0 from Luf, the Hermit Islands at 2 S; 145 E) has dowels to fix the high posts at 71 prow and stern but planks that are entirely sewn, as revealed by a recent X-ray examination at the Museum fUr Volkerkunde, Berlin, in 1980. Most of the museum examples of the mon and other planked boats from the western margin of the Pacific have not been examined, so some potentially available data could be collected.

42. I am indebted to Dr. Ian Glover for discussion of the probable date when metal tools became available, and refer the reader to chapters 7 and 8 of Bell wood (1978) and Chapter 13 of Johnstone (1980). Iron arrived about the same time as bronze. Iron fotming with double bamboo bellows is an ancient technology of Indonesia; iron smelting was carried out in the Philippines when the Spanish arrived. Early accounts show that some Pacific islanders knew the value of iron but had none.

43. In an article in press in the Mariners' Mirror, I argue that the tripod mast, tilted rectangular sail, pulley, metal tools, curved chisel, internal dowels, loss of the lugs on the planks, and probably the quarter rudder mounted on a special support, are all brought from the Indian Ocean to Indonesia by traders after 500 BC. Doran (1981) agrees, with some of these items. The quite different lis-alis and janggolan construction style (Horridge 981, Chapt. 8) could have come from Chinese boatbuilders serving the Dutch in the region of Surabaya in the 19th century.

44. The two-boomed triangular sail pushed up by a movable prop, and held by a stay to the windward outrigger, was called the primitive Oceanic lateen by Haddon & Hornell (1938, Part III.48) but they use the word 'mast' for the prop. In the present article and previously (Horridge 1981, Fig. 12a and plate D) I have called this rig "the protolateen rig", but in future will adopt the term "crane sprit" now introduced by E. Doran in his book "Wangka, Austronesian Canoe Origins", 1981, Texas A & M Univ. Press. It is one of the three Austronesian rigs whereby a heavy mat sail can be pushed up at sea without the use of a pulley (Herridge, in press, Mariners' Mirror).

72 REFERENCES

Bellwood, P., 1978. Man's conquest of the Pacific Collins. . Blair, E.H. & Robertson, J.A., 1903-9. The Philippine Islands 1493-1898. 4 vols (1903) and a series in 55 vols. Cleveland Ohio. USA. Boxer, C.R., 1967. Francisco Vieira de Figueiredo Verhandelingen v.h. Koninklijk lnstituut voor Taal-, Land- en Volkenkunde. 52, 1-118. Nijhoff. 's-Gravenhage, Holland. Dampier, William, 1697. A new voyage 8I'OU1d the world Argonaut Press Edn. 1927, London. Dodd, E., 1972. Polynesian seafaring Nautical Publ. Co. Lymington, England. Doran, E., 1972. Wa, Vinta and Trimaran. J. Polynesian Soc. 81. pp. 144-159. Doran, E., 1981. Wangka. Austronesian Canoe Origins Texas A & M Univ. Press, College Station. Freycinet, Louis C. de Saulees de, 1828. Voyage autour du Monde Atlas, plates 23, 24, 37. Paris. Friederici, G., 1912. Wissenschaftliche Ergebnisse einer amtlichen Forschungsreise nach dem Bismarck Archipel im Jahre 1908. Mitteil~ - den Deutachen Schutzgebieten Erganzungheft 5. publ. Mittler, Berlin. Greenhill, B., 1976. Archaeology of the boat Adam & Charles Black. London. Haddon, A.C., 1937. Canoes of Vol. 2, Melanesia, Queensland and New Guinea. Bishop Museum. Honolulu. Special Publ. 28. Republished 1975. Hornell, James, 1920. The outrigger canoes of Indonesia. Madras Fisheries Bulletin. Vol. 12, Report No. 2, 43-114. Hornell, J., 1935. Constructional parallels in Scandinavian and Oceanic boat construction. Mariner's Mirror 21, 411-427. Hornell, J., 1936a. Canoes of Oceania Vol. I, Polynesia, Fiji and Micronesia. Bishop Museum. Honolulu. Special Publ. 27. Republished 1975. Hornell, J., 1936b. Boat construction in Scandinavia and Oceania. Another parallel in Bote! Tobago. Man, J6, 145-7. Hornell, J., 1946. Water Transport Cambridge University Press. Republished 1970. David & Charles. Newton Abbot. Horridge, G.A., 1978. The design of planked boats of the Moluccas. Monographs and

Reports. ~M Greenwich No. 38. Horridge, G.A., 1979. The Konjo boatbuilders and the Bugis prahus of south Sulawesi.

Monographs and Reports ~M Greenwich No. 40. Horridge, G.A., 1981. The Prahu. Traditional Sailing Boat of Indonesia Oxford University Press. Kuala Lumpur.

73 Jacobs, H. Th. Th. M. (S.J.), 1975. (Ed. and transl.) A treatise on the Moluccas (c.1544), probably the preliminary version of Antonio Galvao's lost "Historia W. Molucas" publ. Jesuit Historical Institute. 00193 Rome. Johnstone, P., 1980. The seacraft of prehistory Routledge & Kegan Paul, London. Leach, E.R., 1937. Boat construction in Botel Tobago. Man 37. 185-7. Manguin, P-Y, 1979. The Southeast Asian trading ship. An historical approach. Proc. I.C.I.O.S. Conf. Univ. of Western Australia, Perth. Australia. August 1979. Markham, A.H., 1880. Voyages and work of John Davis, the navigator Hakluyt Society. Part 50. Page 150. Morga, Antonio de, 1609. Sucesos de las Islas Filipinas Publ. Hakluyt Society 2nd Series. No. 140. 1971. Mussenbroek, S.C.J.W., 1877. Cachelot-visscherij van de Amerikanen in den Ned. lndischen Archipel. Tydachr. My. LLv. Nijverheid Vol. 40, 473-489 and 505-526. Haarlem. Nooteboom, C., 1932. De Boomstamkano in lndonesil! Brill. Leiden. Nooteboom, C., 1951. Aziatische Galeien Overdruk uit de verslagen van het Vierde. International Congress van de Zee. Ostende. July 1951. Publ. Museum of Ethnography & Maritime Musuem "Prins Hendrik" Rotterdam. Paris, F .E., 1843. Essai sur Ia construction navale des peuples extra-Europt{ens publ. Arthus Bertrand, Paris. n.d. (1841-3). Paris, F.E., Admiral, 1882-92. Souvenirs de Marine Republished 1976. Editions des 4 Seigneurs. Grenoble. Rumphius, G.E., 1741. Het Amboinsche Kruid-Boek publ. Amsterdam, Hague, Utrecht. by F. Changuin and others. Schurz, W.L., 1939. The Manila Galleon New York. Scott, W.H., 1980. Boatbuilding and seamanship in classic Philippine society. MS (in press). Spoehr, A., 1980. Protein from the Sea Technological change in Philippine capture fisheries. pp. 224. Ethology Monographs. No. 3. Univ. Pittsburgh, USA. Stellar, KFG and Aebersold, W.E., 1959. Sangirees-Nederlands Woordenboek Vatter, Ernst, 1932. Unbekan1te Bergvolker im tropischen Holland Leipzig: Bibliographisches Institut AG. Wallace, A.R., 1869. The Malay Archipelago MacMillan London (many edns.) Reprinted 1962 Dover, New York. Warington-Smith, H., 1902. ·aoats and boatbuilding in the Malay Peninsula. Journal of the Society of Arts May 16, Vol. 50, 570-586 and pp. 865-867. Weber, M, 1902. lets over Wolvischvangst in den Indischen Archipel. In Rumphius Gedenkboek, p. 89 and also Rep. Siboga Exped. Vol. 3, pp. 31-35 and 146-147. 74 Weber, M., 1923. Die Cetaceen der Siboga-Expedition. Vorkommen und fang der Cetaceem in Indo-Australischen Archipel. Rep. Siboga Exped. Vol. 58. Brill, Leiden. Waldemar, Fred, 1660. View of Macassar. Original at Socie"tl Ge"ographique, Paris, reproduced as Plate III in Boxer 1967. Woodward, D., 1805. Narrative of Capt. David Woodward •• in the island of Celebes 2nd Ed. J. Johnson, 1805. Reprinted 1969, Dawson London. Wright, E.V., 1976. The North Ferriby Boats. National Marit. Mus. Greenwich. Monograph No. 23.

75